Towards defining muscular regions of interest from axial magnetic resonance imaging with anatomical cross-reference: part II - cervical spine musculature

Towards defining muscular regions of interest from axial magnetic resonance imaging with... Background: It has been suggested that the quantification of paravertebral muscle composition and morphology (e.g. size/shape/structure) with magnetic resonance imaging (MRI) has diagnostic, prognostic, and therapeutic potential in contributing to overall musculoskeletal health. If this is to be realised, then consensus towards standardised MRI methods for measuring muscular size/shape/structure are crucial to allow the translation of such measurements towards management of, and hopefully improved health for, those with some musculoskeletal conditions. Following on from an original paper detailing methods for measuring muscles traversing the lumbar spine, we propose new methods based on anatomical cross-reference that strive towards standardising MRI-based quantification of anterior and posterior cervical spine muscle composition. Methods: In this descriptive technical advance paper we expand our methods from the lumbar spine by providing a detailed examination of regional cervical spine muscle morphology, followed by a comprehensive description of the proposed technique defining muscle ROI from axial MRI. Cross-referencing cervical musculature and vertebral anatomy includes an innovative comparison between axial E12 sheet-plastinates derived from cadaveric material to a series of axial MRIs detailing commonly used sequences. These images are shown at different cervical levels to illustrate differences in regional morphology. The method for defining ROI for both anterior (scalenes group, sternocleidomastoid, longus colli, longus capitis) and posterior (multifidus, semispinalis cervicis, semispinalis capitis, splenius capitis) cervical muscles is then described and discussed in relation to existing literature. Results: A series of steps towards standardising the quantification of cervical spine muscle quality are described, with concentration on the measurement of muscle volume and fatty infiltration (MFI). We offer recommendations for imaging parameters that should additionally inform a priori decisions when planning investigations of cervical muscle tissues with MRI. Conclusions: The proposed method provides an option rather than a final position for quantifying cervical spine muscle composition and morphology using MRI. We intend to stimulate discussion towards establishing measurement consensus whereby data-pooling and meaningful comparisons between imaging studies (primarily MRI) investigating cervical muscle quality becomes available and the norm. Keywords: Cervical spine, Paravertebral muscles, Muscle fat infiltration, Magnetic resonance imaging, Region of interest, Manual segmentation * Correspondence: jim.elliott@sydney.edu.au Faculty of Health Sciences, The University of Sydney, Northern Sydney Local Health District, St Leonards, Australia 75 East Street Lidcombe NSW, Sydney 2141, Australia Department of Physical Therapy and Human Movement Sciences, Feinberg School of Medicine, Northwestern University, Chicago, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 2 of 14 Background quantify muscle composition based on differential tissue Magnetic resonance imaging (MRI) has been widely and signal intensities of paravertebral muscle, even the latest, variably utilised to qualify and quantify musculoskeletal time-efficient tools require a degree of manual input for pathology involving a number of soft-tissues in both defining regions of interest (ROI) [3, 5, 6, 38–41]. A traumatic [1–6] and non-traumatic [7, 8] neck disorders. standardised ROI method is arguably most important Such methods have provided convergent [9, 10] and for these studies where it has been speculated that divergent [11–15] evidence around insight into tissue difficulties identifying morphology of both the cervical composition, disease characterisation, response to injury, and lumbar musculature results in poorer repeatability and changes in somatic and nervous structures [6, 38]. With continued improvements in both the potentially due to biological, psychological, and socioen- uptake of, and imaging quality from, MRI technology, an vironmental stresses. Advances in MRI technology have agreed analysis plan utilising a common research raised the number of investigations quantifying skeletal measurement method for the identification of ROIs muscle composition (MFI) and structure (volume, cross- could result in meaningful comparisons with a target sectional area (CSA)), but not without equivocal results towards knowledge transfer and clinical translation of [9]. This variability in findings is likely the result of muscle imaging. Following on from the recent manuscript methodological differences across research groups, detailing a method for determining ROI in the lumbar including variables such as study design, participant spine [42], the purpose of this proposed method is to demographics (trauma vs. non-trauma; sex, sociocul- provide a standardised MRI procedure for measuring tural, age range), measurement techniques, and MR cervical spine muscle composition. The method also parameters used by investigators. serves to initiate and continue discussion on the analysis In order to better understand the influence of muscle of skeletal muscle composition amongst and between the composition and structure on cervical spine health, it is global clinical and scientific communities. imperative that clinical research communities explore and establish common methodologies in order to facili- Method tate standardisation and accurate comparison of data be- Challenges for producing a region of interest of cervical tween studies. Doing so should ultimately result in an muscles using MRI improved understanding of the aetiological features of A number of conventional MRI applications (T1, T2, muscle composition and facilitate an improved prognos- proton-density, Gradient Echo) are available and have tic, diagnostic, and theranostic landscape. been used to qualitatively and quantitatively measure the While data for age-related, degenerative changes of water and fat species of healthy and diseased soft-aqueous tissues (e.g. vertebrae, joints, discs, muscles) of the lum- skeletal muscle tissue [1, 3, 41, 43–49]. Technological bar and cervical spine have been published [16–27], advancements have also produced alternatives that can be studies assessing age-related alterations in paravertebral used to image muscle, such as dual acquisition methods, muscle morphology [19, 28, 29] remain unique to the where frequency is selectively excited to produce a water healthy lumbar spine. Such normative data, to our image [50] and a standard image of fat and water. This, knowledge, does not exist for the cervical spine. While however, produces a challenge when measuring a cross-sectional and longitudinal studies indicate a redundant and anatomically complex set of multi-layered positive relationship between MFI and traumatic neck (and small) muscles in the cervical spine. The challenge is pain (e.g. whiplash associated disorders) [1–3, 5, 6, 30], further compounded by the advent of higher field inconsistent associations are also reported [11–14]. Such scanners (e.g. 3–7 Tesla), where a uniform frequency inconsistencies have not improved our mechanistic un- difference between fat and water content may be difficult, derstanding of changes in muscle composition in both but certainly not impossible, to achieve. traumatic and non-traumatic neck pain. Future works Despite recent technological advances that have must collectively control for what might be considered permitted further insight into muscle composition, the normative age-related changes [19, 29, 31], degenerative mechanisms underlying muscle degeneration and their features of the vertebrae or discs [13, 26, 31–34], and influence on outcomes in neck disorders remain elusive. spinal curvature [35–37]. In addition, the vast majority of symptomatic and asymptomatic population-based studies examining A way forward through standardisation of methodology pathoanatomical features (e.g. the intervertebral disc, In order to facilitate widespread adoption of agreed and ligaments, and the skeletal vertebral column) of the time-efficient techniques for measuring cervical spine cervical spine have used a variety of conventional MRI muscle quality, a standardised, reliable, and replicable sequences [1, 2, 12, 13, 26, 30, 34, 51–55]. Despite the method is urgently required. While there is a general large repository of available works, the data derived from trend toward optimising automated methodologies that these imaging investigations have not revealed a consistent Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 3 of 14 structural lesion(s), or response to said lesion(s), that have The anatomical study we use and recommend for clarified the clinical presentation of traumatic or reference are those detailed in Au et al. [57]. They non-traumatic neck disorders. This has, in our have provided a comprehensive series of labelled axial opinion, created a clinical (and research) impasse MR images from one individual to serve as a that we believe is due partly to the heterogeneous reference atlas of the cervical spine musculature to methods across a number of high quality studies guide clinicians and researchers in the accurate investigating the usefulness of imaging for understanding identification of these muscles on MR imaging. We spinal pathology. Ultimately, the clinical value of imaging have further reinforced by cross-referencing with the findings of spinal pathology and/or muscle degeneration E-12 plastinates that have previously been used to will be realised if such findings predict important assist morphological studies [42, 58]. outcomes or help to identify patients likely to respond to specific interventions (e.g. spinal phenotypes). Anterior muscles Research efforts that focus on the consistent assessment Sternocleidomastoid (SCM) of spinal muscle quality with MRI may improve our The SCM arises from the manubrium and medial clavicle collective biological understanding of traumatic and inferiorly, and angles laterally and posteriorly towards its non-traumatic neck disorders and why some, but not superior attachments at the mastoid process and superior others, recover spontaneously. Accordingly, a robust nuchal line. This superficial muscle is readily identifiable and easily-replicated platform for acquiring, assessing, in cross-section. While the SCM has four portions, [59]as measuring, analysing, and interpreting imaging data they cross and blend, they are not separable in on muscle composition and morphology is needed. cross-section along their length on MRI. The muscle Currently a wide variety of methods are used to has an oval appearance inferiorly, and superiorly describe the composition and morphology of cervical forms a distinctive ‘comma’ shape (Fig. 1). spine muscles (see Table 1 for a non-exhaustive summary). This represents a key challenge for both Scalenus muscles producing consistent regions of interest of cervical Scalenus anterior arises from the scalene tubercle on the spine muscles and allowing comparison between first rib as a thin tendon antero-lateral to the lung and research studies. pleural cavities, and extends superiorly to attach to the anterior tubercles of the C4–6 (and frequently C3) transverse processes. At the level of the first rib the Anatomically defining the muscles of interest subclavian vein passes anterior to scalenus anterior, The muscles spanning the mid-to-lower cervical spine while the subclavian artery passes between scalenus that are typically examined include: multifidus, anterior and medius, visibly separating these two semispinalis cervicis, semispinalis capitis, splenius muscles. At this level scalenus anterior appears capitis, scalenes, levator scapulae, sternocleidomastoid, rounded in cross-section. Scalenus medius arises from and longus capitis and longus colli. We do not describe the first rib posterior to the groove for the subclavian muscles of the suboccipital region (rectus capitis artery and extends superiorly to attach to the posterior major and minor, and the superior and inferior transverse processes of C1–7. obliquus muscles [56]) as it is not possible to accurately measure a clinically useful ROI of the suboccipital Longus capitis muscles from the typically employed transverse images This muscle is largest at C1, and has a flattened appear- used for assessing cervical musculature. This is because ance immediately anterior to the lateral masses on each no suboccipital muscle has a long axis close to side of the midline. Inferiorly, it remains anterior to the perpendicular to the transverse plane, thus making anterior tubercles of the transverse processes, which al- measurement of useful cross-sectional ROI impractical. lows it to be differentiated from longus colli and the sca- Further, fan shaped muscles such as both rectus capitii lenus muscles, particularly scalenus anterior (Fig. 2)[60]. muscles require special consideration in order to validate useful measures, given a single cross-sectional measure- Longus colli ment along the length of either muscle would pose diffi- Longus colli is recognised by its location in the groove culty for determining whole muscle volume. Future formed between the vertebral bodies and transverse pro- work should include developing imaging protocols for cesses of the vertebrae, extending between C1 and T2/3. the suboccipital muscles as they require more nuanced While longus colli is described as having superior, verti- imaging methods and measures with careful consider- cal, and inferior oblique portions, these are based on at- ation around the highest resolution possible within a tachment sites and are not discernible in cross-section reasonable scan time. [61]. The muscle first becomes visible at C2, emerging Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 4 of 14 Table 1 A non-systematic summary of methods across investigations describing cervical spine muscle analysis using magnetic resonance imaging (MRI) Citation Reliability MRI Slice Selection Muscles of Interest ROI Selection Fat Detection Measure Sequence Elliott et al., Inter-rater T1 Axial images aligned parallel to C2–3 disc; MF Manual Quantitative Pixel Fat Infiltration 2006 [1] (0.94) Measured at single slice per level C3-C7; SSCerv Intensity Elliott et al., Intra-rater most cephalad slice of each vertebral SSCap %MFI = (muscle signal)/ 2008 [7] (0.94) body selected SpCap (fat signal)*100 Elliott et al., UT 2009 [78] Elliott et al., 2011 [2] Fernandez Inter-rater T1 Axial images aligned parallel to C2–3 disc; SSCap Manual N/A CSA De Las Penas (0.80–0.98) measured at single slice per level; most SpCap et al., 2007 [79] cephalad slice of each vertebral body selected Elliott et al., Intra-rater T1 Axial images aligned parallel to C2–3 disc; MF Manual N/A CSA 2007 [47] (0.84–0.99) measured at single slice per level C3-C7; SSCerv Inter-rater most cephalad slice of each vertebral SSCap (0.89–0.96) body selected SpCap UT Okada et al., Intra- rater T2 Measurements from a single axial slice MF Manual N/A CSA 2011 [80] (0.90) aligned parallel to each IVD C3–4, C4–5, SSCerv Matsumoto Inter-rater and C5–6 SSCap et al., 2012 [12] (0.844) SpCap Ulbrich et al., Inter-rater STIR Axial images aligned perpendicular to Deep Extensors Manual N/A CSA 2012 [14] (0.79–0.98) the vertebral body in the middle of a All Extensors 20-slice slab. 2 or 3 overlapping slabs SCM used; measurements from single slice per vertebral level C2, C4, and C5 Elliott et al., Inter-rater T1 vs. Dixon Axial images aligned perpendicular to MF Manual Quantitative Pixel Fat Infiltration 2013 [41] for fat- the spinal cord at the C2-C3 IVD; Intensity vs. Fat water sequence measurement from single slice per (F)-Water (W) % MFI = (fat signal)/(fat (0.83–0.99) vertebral level C3-C7 signal + water signal) *100 Elliott et al., From Elliott T1 Axial images aligned parallel to MF Manual Quantitative Pixel Fat Infiltration 2014 [8] et al., 2007 [47] C2–3 IVD; measurements from single SSCerv Intensity for CSA slice crossing IVDs C2-C3 and C5-C6 SSCap %MFI = (fat signal)/(fat Intra- rater SpCap signal + water signal)*100 (0.84–0.99) LCap/LCol Inter-rater SCM CSA (0.89–0.96) Elliott et al., Not reported Dixon Measurements from single slice per MF Manual Fat-Water Fat Infiltration 2015 [3] vertebral level C3-C7; alignment and slice selection not reported %MFI = (fat signal)/(fat signal + water signal)*100 Abbott et al., Intra-rater (0.98) Dixon Measurements averaged over 5 slices MF + SSCerv Manual with Fat-Water Fat Infiltration 2015 [6] Inter-rater (0.93) for each vertebral level C3-C7; Slab (combined) automatic alignment not reported. quartile measure %MFI = (fat signal)/(fat signal + water signal)*100 Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 5 of 14 Table 1 A non-systematic summary of methods across investigations describing cervical spine muscle analysis using magnetic resonance imaging (MRI) (Continued) Citation Reliability MRI Slice Selection Muscles of Interest ROI Selection Fat Detection Measure Sequence Karlsson et al., For muscle fat Dixon Axial images aligned parallel to vertebral MF Manual Fat-Water Fat Infiltration 2016 [5] Intra-rater segments; measurements averaged over (0.81–0.93) 5 slices for each vertebral level C4-C7 %MFI = (fat signal)/(fat Inter-rater signal + water signal)*100 (0.82–0.97) CSA Au et al., 2016 Not reported T1 Axial images aligned parallel to C2–3 IC, IS, LS, LoCap, Manual N/A N/A [57] intervertebral disc; 3D reconstruction LoC, LCap, LCol, MF, LoCap, LoCerv, SSCap, SSCerv, SCM, UT Fortin et al., From [81]: T2 3D multiplanar reconstruction to align MF Manual ROI with Gray-scale threshold Total CSA 2017 [27] Intra-rater images perpendicular to muscle mass; SSCerv semi-automatic technique to calculate Fortin et al., measurements from a single slice per SSCap muscle/fat CSA of fat within Functional CSA (FCSA) 2018 [81] (0.83–0.99) IVD C2–3 through C6–7 SpCap thresholding total muscle CSA; (= fat free area) Inter-rater technique gray-scale range Fatty Infiltration = FCSA/ determined for Total CSA (0.38–0.98) each slice individually Inoue et al., Intra-rater T1 Measured from single slice per level; most MF Manual Lean muscle CSA: ROI Fatty Infiltration = (Fat 2012 [82] (0.85) T2-fat caudal slice of C3 and most cephalad slice drawn on T1-W images CSA)/(Total muscle CSA) suppression of each vertebral body C4-C7 selected; slab not including fat alignment not reported Total muscle CSA: ROI drawn on fat suppression T2-W image including fat Fat CSA = Total CSA – Lean Muscle CSA Mitsutake Intra-rater T1 Measured from single, most cephalad slice MF Manual Quantitative Pixel MFI index = Muscle et al., 2016 0.85–0.94 at level of injury (C4, C5, or C6) Intensity signal/Fat signal [83] Inter-rater (0.84–0.89) Abbott et al., Intra-rater Dixon Axial images aligned parallel to each MF Manual Qualitative grading Fat Infiltration 2017 [67] (0.77–0.88) IVD; measured from 5 slices across each (0, 1, 2) for each 8 MFI Score: Inter-rater vertebral level C4-C7 regions within 0 = no or marginal fat (0.67–0.82) visualized ROI on 1 = light fat fat image 2 = distinct fat Sum of scores Total # of 2’s Choi et al., Inter-rater T1 Axial images aligned parallel to the inferior Flexor Group: Manual N/A CSA 2016 [84] (0.82) end plate of each vertebral body from C4–5 LCap + LCol Normalized Extensor to C7-T1; measured from single slice per Extensor Group: CSA = (Extensor Muscle vertebral level MF + SSCerv CSA)/(Vertebral body CSA) *100 Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 6 of 14 Table 1 A non-systematic summary of methods across investigations describing cervical spine muscle analysis using magnetic resonance imaging (MRI) (Continued) Citation Reliability MRI Slice Selection Muscles of Interest ROI Selection Fat Detection Measure Sequence Cagnie et al., Inter-rater T1 Measured from a single slice aligned LCap Manual Quantitative Pixel Muscle/Fat Index = Muscle 2009 [85] (0.91) parallel IVD at C4-C5 LCol Intensity signal/Fat signal UT LS SpCap SSCerv MF Uthaikup et al., Intra-rater T1 Axial images aligned parallel to the MF Manual Quantitative Pixel Intensity Fat Infiltration 2017 [86] (0.75–0.96) C2–3 IVD; measured from a single SSCap Inter-rater slice at each vertebral level C2-C3 SpCap MFI = Muscle signal/ (0.84–0.99) LCap Fat signal LCol SCM Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 7 of 14 Fig. 1 Axial E12 plastinated section (a) with schematic illustration (b) Fig. 2 Axial E12 plastinated section (a) with schematic illustration and in-phase magnetic resonance image (c) at approximately C2/3 (b) and in-phase magnetic resonance image (c) at approximately identifying musculature at this vertebral level. 1. Longus colli; 2. Longus C5/6 identifying musculature at this vertebral level. Dashed red capitis; 3. Intertransversarii; 4. Levator scapulae; 5. Sternocleidomastoid; (b) and white (c) line indicates an anatomical plane which can 6. Longissimus capitis; 7. Splenius cervicis; 8. Inferior obliquus; 9. Rectus be used as a reference point for identifying some anterior muscles. capitis posterior major; 10. Semispinalis capitis; 11. Splenius capitis; Dashed white line in (c) indicates likely border between multifidus and 12. Trapezius semispinalis cervicis. 1. Sternocleidomastoid; 2. Longus colli; 3. Longus capitis; 4. Scalenus anterior; 5. Scalenus medius; 6. Splenius cervicis; 7. Multifidus / semispinalis cervicis; 8. Semispinalis capitis; 9. Splenius capitis; medial to longus capitis and initially with a more 10. Levator scapulae; 11. Trapezius rounded appearance. Inferior to C7 the muscle thins and moves towards the midline, before attaching to the an- terolateral vertebral bodies. Fascial borders between muscles are only seen in slices between transverse pro- longus colli and the intertransversarii muscles may not cesses (Fig. 1). be readily apparent between any of the cervical levels on MRI. This should not, however, present difficulties as Posterior muscles long as the bony transverse processes are well visualised. Multifidus and rotatores Longus colli remains immediately anterior and medial to Deep against the vertebra, these architecturally complex the bony transverse processes. The intertransversarii muscles fill the space between the spinous and transverse Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 8 of 14 processes. Multifidus is present along the length of the between these two muscles [62], it can be difficult to spine below C2, forming the deepest layer (Figs. 2, 3). identify them as separate entities on both E12s and MRI. Rotatores can be considered together with multifidus in this deep muscle layer, as these muscles are small and do Semispinalis cervicis not form a distinct layer able to be identified in cross- Semispinalis cervicis extends between the spinous section. Together with semispinalis cervicis, multifidus sits processes of C2–5 and the transverse processes of T1- in the paravertebral gutter between the spinous and trans- T5 [63] (Figs. 1, 2, 3). It overlies multifidus along with verse processes. Because of the intimate relationship other cervical-attaching erector spinae (longissimus cervicis, iliocostalis cervicis). The semispinalis cervicis and erector spinae muscles are difficult, if not impos- sible, to adequately distinguish in cross-section. The close approximation, similar alignment and attachments of multifidus, semispinalis and erector spinae fascicles are such that a distinct layer will not always be clear on MRI. In this situation it is reasonable to consider these muscles together as a single group (seen [64] and [6]). Semispinalis capitis Semispinalis capitis is a major muscle of the cervical spine, overlying semispinalis cervicis and forming a large and distinct muscle layer. While semispinalis capitis spans between the occiput and T6–7[63], in cross- section this layer is most apparent between the occiput and C6/7. Below this level this muscle layer becomes less distinct as semispinalis thins and becomes tendinous to- wards the thoracic transverse processes (Figs. 1, 2). Erector spinae Longissimus cervicis extends between the thoracic transverse processes of T1–4 and the C2–6transverse processes, while iliocostalis cervicis passes between the angles of ribs 3–4 and the transverse processes of C4–6 [63]. As noted, erector spinae muscles attaching to the cervical spine are unlikely to be differentiated from semi- spinalis cervicis. Longissimus capitis is more distinct, ex- tending between the mastoid process and the transverse processes of approximately C4-T4 (Fig. 1)[63]. Splenius capitis and cervicis Splenius capitis and cervicis form a single layer and overlie semispinalis capitis. Splenius capitis spans be- tween the C7-T4 spinous processes and the mastoid process / occiput, while splenius cervicis spans between the T3–6 spinous processes and the transverse processes of C1–3[63]. In cross-section, splenius capitis forms a Fig. 3 Axial E12 plastinated section (a) with schematic illustration (b) distinct layer between trapezius and semispinalis capitis. and in-phase magnetic resonance image (c) at approximately C7/T1 Splenius cervicis can be identified between C2–6 on the identifying musculature at this vertebral level. Red box indicates antero-lateral edge of this layer (Figs. 1, 2, 3), as it boundary for Fig. 4. 1. Sternocleidomastoid; 2. Scalenus anterior; 3. Longus colli; 4. Scalenus medius; 5. Iliocostalis cervicis; 6. Multifidus / diverges from splenius capitis towards its cervical attach- semispinalis cervicis; 7. Serratus posterior superior; 8. Splenius capitis ments. Below the level of approximately C5 splenius / cervicis; 9. Levator scapulae; 10. Serratus anterior; 11. cervicis is unlikely to be visibly separate from splenius Rhomboid minor capitis in cross-section. Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 9 of 14 Levator scapulae Levator scapulae have a presence throughout the cervical spine, and its presence is worth noting as one of the larger and more distinctive muscles in the region. It passes from the upper aspect of the medial scapula to the transverse processes of C1–4[63]. In cross-section levator scapulae is well-defined at lower levels, sitting anterior to trapezius and lateral to splenius (Figs. 2, 3). Superiorly, levator scapulae extends towards the trans- verse processes of C1–4 in close relation to the scalenus and longus capitis muscles (Fig. 1). Results Our method provides anatomical reference between MRI imaging and E12 plastinates (derived from cadavers) to advance ROI identification and definition to improve standardised measurement of musculature traversing the cervical spine. The E12 plastinates provide a unique opportunity to detail specific tissues that may be MR invisible, [65] leading to natural disagreement across studies where fat-water separation is a target. To follow, we also include suggestions on operational Fig. 4 Enlarged region of posterior cervical muscles from Fig. 3 (c), characteristics for acquiring MR images. highlighting deep boundary of region of interest (white solid line along lamina). Common mistakes for determining this region of Defining the regions of interest from MRI interest for the transversospinal (TSP) muscles include the boundary Similar to that reported for the lumbar spine, [42]a of multifidus and semispinalis cervicis (white dashed line) or through standard scout image from the sagittal localiser or the fatty infiltrate in multifidus (red dashed line) conventional T2-weighted scan can be used to cross- reference and discern cervical level from axial MR. Users will also find it useful to scroll between the adjacent point for identifying the anterior aspect of all anterior axial slices to accurately landmark anatomical structures muscles apart from the sternocleidomastoid (Fig. 2). when producing ROIs. The method is applicable to studies examining paravertebral ROIs for single (cross- Sternocleidomastoid sectional) or multiple (volumetric) slices. Previous work This definition can be applied along the full extent of from the lumbar spine suggests a randomised approach sternocleidomastoid, from the occiput to approximately for starting with either the left or right side, and/or sep- T2/3. The anatomical boundaries of sternocleidomastoid arate muscles can influence repeatability when creating are straight forward, and tracing should present few ROIs [38, 66]. The same randomised approach is challenges. Some care is needed to trace along the full suggested for the cervical spine. occipital extent at higher levels (Fig. 1). Definitions for ROI measures from MRI for the multi- fidus, semispinalis cervicis, semispinalis capitis, longissi- Scalenus muscles mus capitis, splenius capitis and cervicis, levator This definition is best applied at the C6-T2 levels. The scapulae, longus colli, longus capitis, scalenus and scalenus muscles are best identified at their inferior sternocleidomastoid are included, describing the ana- extent arising from the first rib. Superiorly scalenus tomical borders (cross referenced to Figs. 1, 2, 3). ROI anterior and scalenus medius converge, and may be definitions are detailed with particular reference to cer- difficult to differentiate above the level of C6 on MRI. vical levels C2/3, C5/6, and C7/T1. Technical notes are Differentiation is aided by the angle each muscle ap- also provided where identifying the guided ROI on MRI proaches the cervical transverse processes, as each may be difficult. muscle follows a straight course. Sequentially from anterior to posterior: longus capitis is seen anterior to Anterior muscles the anterior tubercles, scalenus anterior angles to attach It is worth noting that an anatomical plane that passes to the anterior tubercles slightly more laterally, scalenus laterally and posteriorly in an arc from the anterior medius angles between the anterior and posterior tuber- aspect of the vertebral body presents a reliable reference cles, scalenus posterior (if present) angles towards the Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 10 of 14 posterior tubercles, and (above C4) levator scapulae also Semispinalis capitis angles to attach to the posterior tubercles (Fig. 2). This definition is best applied at the occiput-C6 levels. The muscle forms a distinct anatomical layer and can be traced consistent with the anatomy described. As semi- Longus capitis spinalis capitis is clearest at higher levels, difficulties This definition is best applied at the C1–5 levels. Longus identifying this muscle at lower cervical levels would capitis is largest and most distinct superiorly, just below benefit from reviewing and toggling between multiple where it attaches to the basi-occiput. As such, the slices from superior to inferior. As the E12 slices high- muscle is best tracked inferiorly from this point. At its light, it may not be realistic to identify this muscle below superior extent longus capitis has a rounded appearance, approximately C7. which flattens and thins out over the lateral masses of C1. By the level of C2/3 longus capitis is a relatively thin slip immediately anterior to the anterior tubercles of Longissimus capitis transverse processes C3–6. As for the scalenus muscles, This definition is best applied at the C1–4 levels. Longis- identification is aided by identifying the transverse pro- simus capitis is most easily identified as a rounded cesses (in particular the anterior tubercles) and muscle at its superior extent, just below the mastoid remaining posterior to the prevertebral fascia (Fig. 2). process. Towards C4 the muscle flattens, and below ap- proximately C4 it becomes difficult to distinguish from Longus colli other muscles. This definition is best applied at C2-T1 levels. As noted anatomically, longus colli sits in the groove between the vertebral bodies and transverse processes of the verte- Splenius capitis and cervicis brae. Thus, these bony landmarks must be well visua- This definition is best applied at the C1-T3 levels. Sple- lised to accurately identify the muscle. As described for nius capitis is identifiable as a distinct layer located be- multifidus, the ROI should closely follow the bony verte- tween trapezius / sternocleidomastoid and the brae to include fat adjacent to the bone. If the anterior semispinalis capitis. Care is needed around the level of aspect of the transverse processes are not visible, or the mastoid process not to confuse the superior extent slices above and below are not reviewed to clarify the of splenius capitis with sternocleidomastoid or longissi- position of bony landmarks, a ROI for longus colli may mus capitis, which share attachment to the mastoid be inaccurate. process. Just below the mastoid process at the level of C1/2 the muscles appear closely layered from superficial to deep: sternocleidomastoid, splenius capitis (both an Posterior muscles elongated comma shape), and longissimus capitis Multifidus and semispinalis Cervicis (rounded in appearance). Below this level the muscles This definition is best applied at the caudal portion of diverge. Ideally, splenius cervicis will be able to be dis- the C4 vertebral body through the entire T1 vertebral tinguished from splenius capitis at the levels of C2–6 body. With current technology it is generally not pos- (Figs. 1, 2). However, this may not be realistic with sible to consistently delineate between the cervical por- current MRI technology. In this situation, it is reason- tions of the semispinalis cervicis and multifidus on the able to include splenius capitis and cervicis together as a axial slices. While measuring the two muscles independ- single ROI. ently is recommended, they can be combined to form one measure (Figs. 2, 3). As evidenced from the lumbar spine [42], the same approach of approximating the spinous process or lamina is recommended and should Levator scapulae be included within the ROI defining cervical multifidus This definition is best applied at the C2- T1 levels. (Fig. 4). A challenge for both novice and expert clini- While not part of the intrinsic cervical spine muscula- cians remains what to do when creating ROIs between ture, cross-sectional views highlight the presence and the spinous processes. Whether referencing the lumbar size of levator scapulae throughout the cervical spine. [42] or cervical spine, fat commonly overlies the inter- This muscle is largest inferiorly above where it arises spinous space, remains defined, and should be included from the upper part of the medial scapula border, and as when generating ROIs on these slices. Finally, when the such is best tracked superiorly from this point. Care is interspinous ligaments are clearly distinct with a slightly needed to distinguish levator scapulae from serratus irregular and darkened edge, their lateral contour can be anterior as they converge on the scapula (Fig. 3). Atten- followed rather than the spinous process in defining the tion to slices above and below the level of interest will medial border [42]. help resolve their borders. Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 11 of 14 MR imaging - operational parameters [3, 5, 6, 27, 41]. Such an approach has revealed not The type, quality, and output of images acquired from only improved inter- and intra-rater reliability when MR scans are highly influenced by many factors includ- following the spinous process and/or lamina in the ing, but not limited to, user-prescribed parameters. Simi- cervical spine, but also the ability to discriminate be- lar to our previous paper covering the lumbar spine, [42] tween clinical groups [6]. This improved repeatability we endorse consistency in the adoption of MR imaging for defining MF over ES in the lumbar spine has also parameters to facilitate standardised operational proce- been demonstrated [38]. dures that allow intra-study/−institutional comparison and future pooling of results for meta-analyses. The parameters listed here are based on those widely Measures of muscle size and fat utilised in literature (refer to Table 1), and are adapted Measures of muscle size are frequently reported in MRI from those published in a previous paper on ROI for and other imaging-based studies (e.g. ultrasound). In lumbar spine muscles [42]. The parenthetical values both the lumbar and cervical regions, methods employ- given with each parameter are not definitive or unique ing a single cross-sectional MR slice are time efficient to a cervical spine study; rather they are displayed as an for determining muscle size and fat proportion within example of the consistent reporting style we propose. At an ROI. However, a CSA measure from a single-slice a minimum, we believe the following information should should not be taken to constitute a whole muscle size or be reported in all submitted manuscripts: Field strength fat measure [15, 68]. Accordingly, volumetric measures, (e.g. 3 Tesla); sequence type (e.g. 2-point DIXON (3D may be more appropriate [15, 69, 70]. We therefore rec- fast-field echo T1) whole body); repetition time (e.g. TR ommend a multi-slice approach that derives muscle size 4.2 ms); echo time (e.g. TE 1.2 and 3.1 ms); flip angle (e. and fat content based on a three-dimensional volume g. 5°); field of view (e.g. FOV 560 × 352 mm); acquired across the levels of interest. In going forward, such voxel dimensions (e.g. 2.0 × 2.0 × 4.0 mm); reconstructed measures should be accurately categorised as a 3- voxel dimensions (e.g. 1.0 × 1.0 × 2.0 mm); bandwidth (e.g. dimensional volume of the entire muscle as 3D acquisi- 240 Hz/Px), acquisition time (e.g. TA 5 min 22 s) and slice tion methods with MRI have evolved and are not as thickness (e.g. 4.0 mm). Additionally, the description sensitive to the radio frequency slice profile as is 2D should include axial slice alignment (e.g. aligned parallel imaging [15]. to C2–3 intervertebral disc), slice selection (e.g. measure- It is of course acknowledged that acquiring such data ments taken at most cephalad slice per vertebral level), with both semi-automated or automated programmes and subject body position including any support materials for both the lumbar [42] and cervical spines is time- that may influence cervical spine posture/curvature (e.g. consuming. However, with the evolution of higher- subjects positioned supine with arms at sides and 2 in. resolution imaging techniques a more time-efficient foam cushion under head). capture of cervical muscle volumes from a single verte- bral level may correspond to a representative marker of Discussion MFI across the entire cervical column. While this has A foundational edict for defining lumbar paravertebral been demonstrated in the healthy lumbar spine [29] ROI’s from MRI studies has previously been published where the fat content at L4 best represents that of the [42]. Here, we expand the previous methods [57] for the entire lumbar region, future research should continue to cervical pre- and para-vertebral muscles using a number systematically include the entire cervical spine in healthy of MRI and E12 sheet plastinate illustrations of vertebral and symptomatic cohorts to build a stronger body of morphology with the aim of standardising muscle ROI evidence regarding age-aggregated cervical paravertebral definitions. The E12 plastinates provide a unique oppor- muscle composition. tunity to detail specific tissues that may be MR invisible, Another issue with longitudinal designs, where muscle [65] leading to natural disagreement across studies measures are produced over time, remains a general lack where fat-water separation is a target. Also unique to of reporting on how the MRI slices are aligned in plane. this work is the included suggestions on operational A failure to do so could potentially result in registration characteristics for acquiring MR images. discrepancies depending how the angle through each Similar to the proposed approach in the lumbar spine, muscle was performed. Using some standard anatomical [42] we consider that if fat is occupying space deep to reference (e.g. vertebral bone) that is not expected to ap- the epimysial sheath and close to the spinous processes, preciably change over time could control for this. Errors laminae, zygapophyseal joints, it has a potential bio- of this type can be further minimised by reporting mechanical consequence on muscle function, [6, 67] and muscle volume over the full length of the muscle (from should be included in the ROI (Fig. 4). We base this de- origin to insertion), as suggested above, rather than a cision in part on previous work in the cervical spine single-slice CSA. Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 12 of 14 Measures of muscle fat with MRI expanded, and refined and ultimately result in an estab- The demonstration of neck muscle fatty infiltrates on lished common methodology towards facilitating con- T1- weighted imaging in acute [2, 3] and chronic trau- sistent and accurate definitions of lumbar, cervical, and matic neck pain [1, 5, 8, 30] has been reported in cross- upper/lower limb muscle ROIs on axial imaging, sectional and longitudinal fashion and across three particularly MRI. countries (Australia, [2] Sweden, [5] and the United Abbreviations States [3, 6]). Such findings are not present to the same CSA: Cross-sectional area; FCSA: Functional cross-sectional area; magnitude for those with chronic idiopathic neck pain FSPGR: Fast-spoiled gradient echo; IC: Iliocostalis cervicis; IP: In-phase (water); IS: Interspinalis cervicis; IVD: Intervertebral disc; LCap: Longus [7] and it has been postulated that these muscle changes capitis; LCol: Longus colli; LoCap: Longissimus capitis; LoCerv: Longissimus represent one neurophysiologic basis for the transition cervicis; LS: Levator scapulae; MF: Multifidus; MFI: Muscle fat infiltration; to chronic pain in this population [71]. A variety of OP: Opposed-phase (Fat); SCap: Spinalis capitis; SCerv: Spinalis cervicis; SCM: Sternocleidomastoid; SI: Signal intensity; SpCap: Splenius capitis; newer and more rapid high resolution MRI techniques SSCap: Semispinalis capitis; SSCerv: Semispinalis cervicis; UT: Upper trapezius (3D Fat/Water Separation and Proton-Density Fat Frac- tion, Fat suppression) [65, 72–77] and analyses (FCSA/ Availability of data and materials CSA, Fat Signal Fraction, MFI %) could help better visu- All the data supporting the findings are contained within the manuscript. alise and quantify physiologic changes at the level of the Authors’ contributions muscle cell or other disease processes when compared JME and RJC conceived the study, while RJC, JC, EK, RA, and JME each made to other conventional clinical imaging sequences (e.g. substantial contributions to its design. RJC, JC, EK, RA, and JME have been T1- and T2-weighted). However, such variety across involved in drafting the manuscript or revising it critically for important intellectual content and each given final approval of the version to be methods and techniques also complicates comparison published. RJC, JC, EK, RA, and JME agree to be accountable for all aspects of among studies. Accordingly, we call for all authors to the work in ensuring that questions related to the accuracy or integrity of clearly detail their fat infiltration measurements to any part of the work are appropriately investigated and resolved.All authors read and approved the final manuscript. ensure that future pooling of data efforts is possible. Fur- ther, with the number of proprietary semi-automated or Ethics approval and consent to participate automated methods appearing in the literature, and of No ethics approval was required to undertake this descriptive methodological study. The preparation of the E-12 plastinated sections took which descriptions are limited due to commercial sensi- place in accordance with the New Zealand Human Tissue Act (2008) and tivity, we contend it will be helpful for authors to in- University of Otago institutional guidelines. clude enough technical detail for comparisons to the fundamental literature to be made. Consent for publication Approval to use images of the E-plastinated sections was granted by Depart- ment of Anatomy, University of Otago. All MRIs were derived from the same Participant positioning informed and consenting adult subject. It is our recommendation that participants should lie su- Competing interests pine inside the magnet with a foam pad under their Authors RJC, JC, EK, and RA have no disclosures to declare. In unrelated knees and foam padding placed on the right and left of activities, JME is principal investigator on NIH grant [HD079076-01A1; 09/ the head to minimise head movement. A neutral pos- 2014–05/2019]. ition, visually determined by ensuring that a horizontal position of the forehead to the chin is parallel to the Publisher’sNote MRI table, is also recommended. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Conclusion Author details We follow on from, and have expanded, an original Faculty of Health Sciences, The University of Sydney, Northern Sydney Local Health District, St Leonards, Australia 75 East Street Lidcombe NSW, Sydney paper of manually defining ROIs of lumbar spine mus- 2141, Australia. Department of Physical Therapy and Human Movement culature [42] to now include the cervical muscles. While Sciences, Feinberg School of Medicine, Northwestern University, Chicago, the method aims to permit accurate and reliable com- USA. Honorary Fellow School of Health and Rehabilitation Sciences, The University of Queensland, St. Lucia, Australia. Centre for Early Learning in parison of cervical muscle quality between studies in Medicine, Otago Medical School, University of Otago, Dunedin, New Zealand. (and beyond) this field, we further suggest journals adopt School of Physiotherapy, University of Otago, Dunedin, New Zealand. a more robust reporting of imaging parameters used to Faculty of Health Sciences, Curtin University, Perth, Australia. assist consistency and allow accurate comparison be- Received: 14 February 2018 Accepted: 4 May 2018 tween studies. It is imperative to note that we are cognisant the application methods are not definitive end-points on References 1. Elliott J, Jull G, Noteboom JT, Darnell R, Galloway G, Gibbon WW. Fatty ‘how to’ and that there is potential for much repetition infiltration in the cervical extensor muscles in persistent whiplash-associated across body regions. Rather, we hope that with time, and disorders: a magnetic resonance imaging analysis. Spine (Phila Pa 1976). new research findings, these methods will be modified, 2006;31(22):E847–55. Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 13 of 14 2. Elliott J, Pedler A, Kenardy J, Galloway G, Jull G, Sterling M. The temporal 22. Hancock MJ, Kjaer P, Kent P, Jensen RK, Jensen T. Is the number of different development of fatty infiltrates in the neck muscles following whiplash MRI findings more strongly associated with low back painthan single MRI injury: an association with pain and posttraumatic stress. PLoS One. 2011; findings? Spine. 2017;42(17):1283–8. 6(6):e21194. 23. Panagopoulos J, Hush J, Steffens D, Hancock M. Do MRI findings change 3. Elliott JM, Courtney DM, Rademaker A, Pinto D, Sterling MM, Parrish T. The over a period of up to one year in patients with low back pain and/or rapid and progressive degeneration of the cervical multifidus in whiplash: a sciatica? A Systematic Review. Spine. 2017;42(7):504–12. MRI study of fatty infiltration. Spine (Phila Pa 1976). 2015;40(12):E694–700. 24. Steffens D, Hancock MJ, Maher CG, Williams C, Jensen TS, Latimer J. Does 4. Elliott JM, Pedler AR, Theodoros D, Jull GA. Magnetic resonance imaging magnetic resonance imaging predict future low back pain? A systematic changes in the size and shape of the oropharynx following acute whiplash review. Eur J Pain. 2014;18(6):755–65. injury. J Orthop Sports Phys Ther. 2012;42(11):912–8. 25. Wan Q, Lin C, Li X, Zeng W, Ma C. MRI assessment of paraspinal muscles in 5. Karlsson A, Dahlqvist Leinhard O, West J, Romu T, Aslund U, Smedby O, patients with acute and chronic unilateral low back pain. Br J Radiol. 2015; Zsigmond P, Peolsson A. An investigation of fat infiltration of the multifidus 88(1053):20140546. muscle in patients with severe neck symptoms associated with chronic 26. Nakashima H, Yukawa Y, Suda K, Yamagata M, Ueta T, Kato F. Abnormal whiplash associated disorder. J Orthop Sports Phys Ther. 2016;46(10):886–93. findings on magnetic resonance images of the cervical spines in 1211 6. Abbott R, Pedler A, Sterling M, Hides J, Murphey T, Hoggarth M, Elliott J. asymptomatic subjects. Spine (Phila Pa 1976). 2015;40(6):392–8. The geography of fatty infiltrates within the cervical multifidus and 27. Fortin M, Dobrescu O, Courtemanche M, Sparrey CJ, Santaguida C, Fehlings semispinalis Cervicis in individuals with chronic whiplash-associated MG, Weber M. Association between paraspinal muscle morphology, clinical disorders. J Orthop Sports Phys Ther. 2015;45(4):281–8. symptoms and functional status in patients with degenerative cervical 7. Elliott J, Sterling M, Noteboom JT, Darnell R, Galloway G, Jull G. Fatty myelopathy. Spine. 2017;42(4):232–9. infiltrate in the cervical extensor muscles is not a feature of chronic, 28. Amabile C, Moal B, Chtara OA, Pillet H, Raya JG, Iannessi A, Skalli W, Lafage V, insidious-onset neck pain. Clin Radiol. 2008;63(6):681–7. Bronsard N: Estimation of spinopelvic muscles' volumes in young asymptomatic 8. Elliott JM, Pedler AR, Jull GA, Van Wyk L, Galloway GG, O’Leary S. Differential subjects: a quantitative analysis. Surgical and radiologic anatomy: SRA. 2016;39(4): changes in muscle composition exist in traumatic and non-traumatic neck 393–403. pain. Spine (Phila Pa 1976). 2014;39(1):39–47. 29. Crawford R, Filli L, Elliott J, Nanz D, Fischer M, Marcon M, Ulbrich E. Age- 9. De Pauw R, Coppieters I, Kregel J, De Meulemeester K, Danneels L, Cagnie and level-dependence of fatty infiltration in lumbar paravertebral muscles of B. Does muscle morphology change in chronic neck pain patients? - a healthy volunteers. Am J Neuroradiol. 2016;37(4):742–8. systematic review. Man Ther. 2016;22:42–9. 30. Elliott JM, O'Leary S, Sterling M, Hendrikz J, Pedler A, Jull G. Magnetic 10. Thakar S, Mohan D, Furtado SV, Sai Kiran NA, Dadlani R, Aryan S, Rao AS, resonance imaging findings of fatty infiltrate in the cervical flexors in Hegde AS. Paraspinal muscle morphometry in cervical spondylotic chronic whiplash. Spine (Phila Pa 1976). 2010;35(9):948–54. myelopathy and its implications in clinicoradiological outcomes following 31. Bhadresha A, Lawrence OJ, McCarthy MJ. A comparison of magnetic central corpectomy: clinical article. J Neurosurg Spine. 2014;21(2):223–30. resonance imaging muscle fat content in the lumbar Paraspinal muscles with 11. Anderson SE, Boesch C, Zimmermann H, Busato A, Hodler J, Bingisser R, patient-reported outcome measures in patients with lumbar degenerative disk Ulbrich EJ, Nidecker A, Buitrago-Tellez CH, Bonel HM, et al. Are there cervical disease and focal disk prolapse. Glob Spine J. 2016;6(4):401–10. spine findings at MR imaging that are specific to acute symptomatic 32. Teichtahl AJ, Urquhart DM, Wang Y, Wluka AE, Wijethilake P, O'Sullivan R, whiplash injury? A prospective controlled study with four experienced Cicuttini FM. Fat infiltration of paraspinal muscles is associated with low blinded readers. Radiology. 2012;262(2):567–75. back pain, disability, and structural abnormalities in community-based 12. Matsumoto M, Ichihara D, Okada E, Chiba K, Toyama Y, Fujiwara H, adults. Spine J. 2015;15(7):1593–601. Momoshima S, Nishiwaki Y, Takahata T. Cross-sectional area of the posterior 33. Kalichman L, Hodges P, Li L, Guermazi A, Hunter DJ. Changes in paraspinal extensor muscles of the cervical spine in whiplash injury patients versus muscles and their association with low back pain and spinal degeneration: healthy volunteers - 10year follow-up MR study. Injury. 2012;43(6):912–6. CT study. Eur Spine J. 2010;19(7):1136–44. 13. Matsumoto M, Okada E, Ichihara D, Chiba K, Toyama Y, Fujiwara H, 34. Matsumoto M, Ichihara D, Okada E, Toyama Y, Fujiwara H, Momoshima S, Momoshima S, Nishiwaki Y, Hashimoto T, Inoue T, et al. Prospective ten-year Nishiwaki Y, Takahata T. Modic changes of the cervical spine in patients with follow-up study comparing patients with whiplash-associated disorders and whiplash injury: a prospective 11-year follow-up study. Injury. 2013;44(6):819–24. asymptomatic subjects using magnetic resonance imaging. Spine (Phila Pa 35. Meakin JR, Fulford J, Seymour R, Welsman JR, Knapp KM. The relationship 1976). 2010;35(18):1684–90. between sagittal curvature and extensor muscle volume in the lumbar 14. Ulbrich EJ, Aeberhard R, Wetli S, Busato A, Boesch C, Zimmermann H, spine. J Anat. 2013;222(6):608–14. Hodler J, Anderson SE, Sturzenegger M. Cervical muscle area measurements 36. Pezolato A, de Vasconcelos EE, Defino HL, Nogueira-Barbosa MH. Fat in whiplash patients: acute, 3, and 6 months of follow-up. J Magn Reson infiltration in the lumbar multifidus and erector spinae muscles in subjects Imaging. 2012;36(6):1413–20. with sway-back posture. Eur Spine J. 2012;21(11):2158–64. 15. Elliott JM, Kerry R, Flynn T, Parrish T. Content not quantity is a better 37. Johansson MP, Baann Liane MS, Bendix T, Kasch H, Kongsted A. Does measure of muscle degeneration in whiplash. Man Ther. 2013;18(6):578–82. cervical kyphosis relate to symptoms following whiplash injury. Man Ther. 16. Brinjikji W, Luetmer PH, Comstock B, Bresnahan BW, Chen LE, Deyo RA, 2011;16(4):378–83. Halabi S, Turner JA, Avins AL, James K, et al. Systematic literature review of 38. Mhuiris AN, Volken T, Elliott JM, Hoggarth M, Samartzis D, Crawford RJ. imaging features of spinal degeneration in asymptomatic populations. Am J Reliability of quantifying the spatial distribution of fatty infiltration in lumbar Neuroradiol. 2015;36(4):811–6. paravertebral muscles using a new segmentation method for T1-weighted 17. Crawford RJ, Volken T, Valentin S, Melloh M, Elliott J. Rate of lumbar MRI. BMC Musculoskelet Disord. 2016;17(1):234. paravertebral muscle fat infiltration versus spinal degeneration in 39. Putzier M, Hartwig T, Hoff EK, Streitparth F, Strube P. Minimally invasive TLIF asymptomatic populations: an age- aggregated cross-sectional simulation leads to increased muscle sparing of the multifidus muscle but not the study. BMC Scoliosis Spinal Disord. 2016;11(1):21. longissimus muscle compared with conventional PLIF-a prospective 18. Fortin M, Yuan Y, Battie MC. Factors associated with paraspinal muscle randomized clinical trial. Spine Journal. 2016;16(7):811–9. asymmetry in size and composition in a general population sample of men. 40. Valenzuela W, Ferguson SJ, Ignasiak D, Diserens G, Vermathen P, Boesch C, Phys Ther. 2013;93(11):1540–50. Reyes M. Correction tool for active shape model based lumbar muscle 19. Valentin S, Licka T, Elliott J. Age and side-related morphometric MRI segmentation. PLoS One. 2015;2015:3033–6. evaluation of trunk muscles in people without back pain. Man Ther. 2015; 41. Elliott JM, Walton DM, Rademaker A, Parrish T. Quantification of cervical 20(1):90–5. spine muscle fat: a comparison between T1-weighted and multi-echo 20. Valentin S, Licka TF, Elliott J. MRI-determined lumbar muscle morphometry gradient echo imaging using a variable projection algorithm (VARPRO). BMC in man and sheep: potential biomechanical implications for ovine model to Med Imaging. 2013;11:13–30. https://doi.org/10.1186/1471-2342-13-30. human spine translation. J Anat. 2015;227(4):506–13. 42. Crawford RJ, Cornwall J, Abbott R, Elliott J. Manually defining regions of 21. Hancock M, Maher C, Macaskill P, Latimer J, Kos W, Pik J. MRI findings are interest when quantifying paravertebral muscles fatty infiltration from axial more common in selected patients with acute low back pain than controls? magnetic resonance imaging: a proposed method for the lumbar spine Eur Spine J. 2012;21(2):240–6. with anatomical cross-reference. BMC Musculoskelet Disord. 2017:18(25). Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 14 of 14 43. Fleckenstein JL, Watamull D, Conner KE, Ezaki M, Greenlee RG Jr, Bryan WW, 66. Valentin S, Yeates TD, Licka T, Elliott J. Inter-rater reliability of trunk muscle Chason DP, Parkey RW, Peshock RM, Purdy PD. Denervated human skeletal morphometric analysis. J Back Musculoskeletal Rehabil. 2015;28(1):181–90. muscle: MR imaging evaluation. Radiology. 1993;187(1):213–8. 67. Abbott R, Peolsson A, West J, Elliott JM, Aslund U, Karlsson A, Dahlqvist 44. Fritz RC, Domroese ME, Carter GT. Physiological and anatomical basis of Leinhard O: The qualitative grading of muscle fat infiltration in whiplash muscle magnetic resonance imaging. Phys Med Rehabil Clin N Am. 2005; using fat/water magnetic resonance imaging. Spine J 2017, Sep 5. pii: 16(4):1033–51. x S1529–9430(17)30907–5. https://doi.org/10.1016/j.spinee.2017.08.233. [Epub ahead of print]. 45. Wokke BH, Bos C, Reijnierse M, van Rijswijk CS, Eggers H, Webb A, Verschuuren JJ, Kan HE. Comparison of Dixon and T1-weighted MR 68. Elliott JM, Pedler AR, Jull GA, Van Wyk L, Galloway GG, O'Leary SP. Differential changes in muscle composition exist in traumatic and methods to assess the degree of fat infiltration in duchenne muscular nontraumatic neck pain. Spine. 2014;39(1):39–47. dystrophy patients. J Magn Reson Imaging. 2013;38(3):619–24. 69. Boom HP, van Spronsen PH, van Ginkel FC, van Schijndel RA, Castelijns JA, 46. Elliott JM, Galloway GJ, Jull GA, Noteboom JT, Centeno CJ, Gibbon WW. Tuinzing DB. A comparison of human jaw muscle cross-sectional area and volume Magnetic resonance imaging analysis of the upper cervical spine extensor in long- and short-face subjects, using MRI. Arch Oral Biol. 2008;53(3):273–81. musculature in an asymptomatic cohort: an index of fat within muscle. Clin 70. Abbott R, Pedler A, Sterling M, Hides J, Murphey T, Hoggarth M, Radiol. 2005;60(3):355–63. Elliott J. The geography of fatty infiltrates within the cervical 47. Elliott JM, Jull GA, Noteboom JT, Durbridge GL, Gibbon WW. Magnetic multifidus and semispinalis cervicis in individuals with chronic resonance imaging study of cross-sectional area of the cervical extensor whiplash-associated disorders. J Orthop Sports Phys Ther. 2015;45(4):8. musculature in an asymptomatic cohort. Clin Anat. 2007;20(1):35–40. 71. Elliott J. Are there implications for morphological changes in neck muscles after 48. Sinclair CD, Morrow JM, Miranda MA, Davagnanam I, Cowley PC, Mehta H, whiplash injury? Spine (Phila Pa 1976). 2011;1(36(25 Suppl)):S205–10. Review Hanna MG, Koltzenburg M, Yousry TA, Reilly MM, et al. Skeletal muscle MRI 72. Bley TA, Wieben O, Francois CJ, Brittain JH, Reeder SB. Fat and water magnetisation transfer ratio reflects clinical severity in peripheral magnetic resonance imaging. J Magn Reson Imaging. 2010;31(1):4–18. neuropathies. J Neurol Neurosurg Psychiatry. 2012;83(1):29–32. 73. Reeder SB, McKenzie CA, Pineda AR, Yu H, Shimakawa A, Brau AC, 49. Sinclair CD, Morrow JM, Janiczek RL, Evans MR, Rawah E, Shah S, Hanna MG, Hargreaves BA, Gold GE, Brittain JH. Water-fat separation with IDEAL Reilly MM, Yousry TA, Thornton J. Stability and sensitivity of water T2 gradient-echo imaging. J Magn Reson Imaging. 2007;25(3):644–52. obtained with IDEAL-CPMG in healthy and fat-infiltrated skeletal muscle. 74. Costa DN, Pedrosa I, McKenzie C, Reeder SB, Rofsky NM. Body MRI using NMR Biomed. 2016;29(12):1800–12. IDEAL. Am J Roentgenol. 2008;190(4):1076–84. 50. Haase A, Frahm J, Hänicke W, Matthaei D. 1H NMR chemical shift selective 75. Gerdes CM, Kijowski R, Reeder SB. IDEAL imaging of the musculoskeletal (CHESS) imaging. Phys Med Biol. 1985;30:341–4. system: robust water fat separation for uniform fat suppression, marrow 51. Kaale BR, Krakenes J, Albrektsen G, Wester K. Whiplash-associated disorders evaluation, and cartilage imaging. AJR Am J Roentgenol. 2007;189(5):W284–91. impairment rating: neck disability index score according to severity of MRI 76. Romu T, Dahlqvist Leinhard O, Dahlström N, Borga M. Robust water fat findings of ligaments and membranes in the upper cervical spine. J separated dual-Echo MRI by phase-sensitive reconstruction. Magn Reson Neurotrauma. 2005;22(4):466–75. Med. 2017;78(3):1208–16. 52. Krakenes J, Kaale BR. Magnetic resonance imaging assessment of 77. Gerdle B, Forsgren MF, Bengtsson A, Leinhard OD, Soren B, Karlsson A, craniovertebral ligaments and membranes after whiplash trauma. Spine. Brandejsky V, Lund E, Lundberg P. Decreased muscle concentrations of ATP 2006;31(24):2820–6. and PCR in the quadriceps muscle of fibromyalgia patients - a (31) P-MRS 53. Krakenes J, Kaale BR, Moen G, Nordli H, Gilhus NE, Rorvik J. MRI assessment study. Eur J Pain. 2013;17(8):1205–15. of the alar ligaments in the late stage of whiplash injury–a study of 78. Elliott J, Sterling M, Noteboom JT, Treleaven J, Galloway G, Jull G. The structural abnormalities and observer agreement. Neuroradiology. 2002; clinical presentation of chronic whiplash and the relationship to findings of 44(7):617–24. MRI fatty infiltrates in the cervical extensor musculature: a preliminary 54. Myran R, Kvistad KA, Nygaard OP, Andresen H, Folvik M, Zwart JA. Magnetic investigation. Eur Spine J. 2009;18(9):1371–8. resonance imaging assessment of the alar ligaments in whiplash injuries: a 79. Fernandez-de-Las-Penas C, Bueno A, Ferrando J, Elliott JM, Cuadrado ML, Pareja case-control study. Spine. 2008;33(18):2012–6. JA. Magnetic resonance imaging study of the morphometry of cervical extensor 55. Ronnen HR, de Korte PJ, Brink PR, van der Bijl HJ, Tonino AJ, Franke C. Acute muscles in chronic tension-type headache. Cephalalgia. 2007;27(4):355–62. whiplash injury: is there a role for MR imaging?–a prospective study of 100 80. Okada E, Matsumoto M, Ichihara D, Chiba K, Toyama Y, Fujiwara H, Momoshima patients. Radiology. 1996;201(1):93–6. S, Nishiwaki Y, Takahata T. Cross-sectional area of posterior extensor muscles of 56. Cornwall J, Farrell SF, Sheard P. Fibre types of human suboccipital muscles. the cervical spine in asymptomatic subjects: a 10-year longitudinal magnetic Eur J Anat. 2016;20(1):31–6. resonance imaging study. Eur Spine J. 2011;20(9):1567–73. 57. Au J, Perriman DM, Pickering MR, Buirski G, Smith PN, Webb AL. Magnetic 81. Fortin M, Dobrescu O, Jarzem P, Ouellet J, Weber MH. Quantitative resonance imaging atlas of the cervical spine musculature. Clin Anat. 2016; magnetic resonance imaging analysis of the cervical spine extensor 29(5):643–59. muscles: Intrarater and interrater reliability of a novice and an experienced 58. Farrell SF, Osmotherly PG, Cornwall J, Sterling M, Rivett D. Cervical spine rater. Asian Spine J. 2018;12(1):94–102. meniscoids: an update on their morphological characteristics and potential 82. Inoue H, Montgomery S, Aghdasi B, Tan Y, Tian H, Jian X, Terrell R, Singh V, clinical significance. Eur Spine J. 2017;26(4):939–47. Wang J. Analysis of relationship between Paraspinal muscle fatty 59. Kennedy E, Albert M, Nicholson H. The fascicular anatomy and peak force degeneration and cervical spine motion using kinetic magnetic resonance capabilities of the sternocleidomastoid muscle. Surg Radiol Anat. 2017;39(6): imaging. Glob Spine J. 2012;2(1):33–8. 629–45. 83. Mitsutake T, Sakamoto M, Chyuda Y, Oka S, Hirata H, Matsuo T, Oishi T, 60. Cornwall J, Kennedy E. Fiber types of the anterior and lateral cervical Horikawa E. Greater cervical muscle fat infiltration evaluated by magnetic muscles in elderly males. Eur Spine J. 2015;24(9):1986–91. resonance imaging is associated with poor postural stability in patients with 61. Miller A, Woodley SJ, Cornwall J. Fibre type composition of female longus cervical Spondylotic radiculopathy. Spine. 2016;41:1. capitis and longus colli muscles. Anat Sci Int. 2016;91(2):163–8. 84. Choi MK, Kim SB, Park CK, Lee SH, Jo DJ. Relation of deep Paraspinal 62. Cornwall J, Stringer MD, Duxson M. Functional morphology of the thoracolumbar Muscles' cross-sectional area of the cervical spine and bone Union in transversospinal muscles. Spine (Phila Pa 1976). 2011;36(16):E1053–61. Anterior Cervical Decompression and Fusion: a retrospective study. World 63. Standring S, Anand N, Birch R, Collins P, Crossman AR, Gleeson M, et al: Neurosurg. 2016;96:91–100. Gray's anatomy : the anatomical basis of clinical practice., 41st edn. New 85. Cagnie B, Barbe T, Vandemaele P, Achten E, Cambier D, Danneels L. MRI York: Elsevier; 2016. analysis of muscle/fat index of the superficial and deep neck muscles in an 64. Smith AC, Parrish TB, Hoggarth MA, McPherson JG, Tysseling VM, asymptomatic cohort. Eur Spine J. 2009;18(5):704–9. Wasielewski M, Kim H, Hornby TG, Elliott J. Potential associations between 86. Uthaikhup S, Assapun J, Kothan S, Watcharasaksilp K, Elliott JM. Structural chronic whiplash and incomplete spinal cord injury. Spinal Cord Ser Cases. changes of the cervical muscles in elder women with cervicogenic 2015;2015:15024. https://doi.org/10.1038/scsandc.2015.24. headache. Musculoskelet Sci Pract. 2017;29:1–6. 65. Reeder SB, Hu HH, Sirlin CB. Proton density fat-fraction: a standardized mr-based biomarker of tissue fat concentration. J Magn Reson Imaging. 2012;36(5):1011–4. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Musculoskeletal Disorders Springer Journals

Towards defining muscular regions of interest from axial magnetic resonance imaging with anatomical cross-reference: part II - cervical spine musculature

Free
14 pages
Loading next page...
 
/lp/springer_journal/towards-defining-muscular-regions-of-interest-from-axial-magnetic-KxHavotWiV
Publisher
BioMed Central
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Orthopedics; Rehabilitation; Rheumatology; Sports Medicine; Internal Medicine; Epidemiology
eISSN
1471-2474
D.O.I.
10.1186/s12891-018-2074-y
Publisher site
See Article on Publisher Site

Abstract

Background: It has been suggested that the quantification of paravertebral muscle composition and morphology (e.g. size/shape/structure) with magnetic resonance imaging (MRI) has diagnostic, prognostic, and therapeutic potential in contributing to overall musculoskeletal health. If this is to be realised, then consensus towards standardised MRI methods for measuring muscular size/shape/structure are crucial to allow the translation of such measurements towards management of, and hopefully improved health for, those with some musculoskeletal conditions. Following on from an original paper detailing methods for measuring muscles traversing the lumbar spine, we propose new methods based on anatomical cross-reference that strive towards standardising MRI-based quantification of anterior and posterior cervical spine muscle composition. Methods: In this descriptive technical advance paper we expand our methods from the lumbar spine by providing a detailed examination of regional cervical spine muscle morphology, followed by a comprehensive description of the proposed technique defining muscle ROI from axial MRI. Cross-referencing cervical musculature and vertebral anatomy includes an innovative comparison between axial E12 sheet-plastinates derived from cadaveric material to a series of axial MRIs detailing commonly used sequences. These images are shown at different cervical levels to illustrate differences in regional morphology. The method for defining ROI for both anterior (scalenes group, sternocleidomastoid, longus colli, longus capitis) and posterior (multifidus, semispinalis cervicis, semispinalis capitis, splenius capitis) cervical muscles is then described and discussed in relation to existing literature. Results: A series of steps towards standardising the quantification of cervical spine muscle quality are described, with concentration on the measurement of muscle volume and fatty infiltration (MFI). We offer recommendations for imaging parameters that should additionally inform a priori decisions when planning investigations of cervical muscle tissues with MRI. Conclusions: The proposed method provides an option rather than a final position for quantifying cervical spine muscle composition and morphology using MRI. We intend to stimulate discussion towards establishing measurement consensus whereby data-pooling and meaningful comparisons between imaging studies (primarily MRI) investigating cervical muscle quality becomes available and the norm. Keywords: Cervical spine, Paravertebral muscles, Muscle fat infiltration, Magnetic resonance imaging, Region of interest, Manual segmentation * Correspondence: jim.elliott@sydney.edu.au Faculty of Health Sciences, The University of Sydney, Northern Sydney Local Health District, St Leonards, Australia 75 East Street Lidcombe NSW, Sydney 2141, Australia Department of Physical Therapy and Human Movement Sciences, Feinberg School of Medicine, Northwestern University, Chicago, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 2 of 14 Background quantify muscle composition based on differential tissue Magnetic resonance imaging (MRI) has been widely and signal intensities of paravertebral muscle, even the latest, variably utilised to qualify and quantify musculoskeletal time-efficient tools require a degree of manual input for pathology involving a number of soft-tissues in both defining regions of interest (ROI) [3, 5, 6, 38–41]. A traumatic [1–6] and non-traumatic [7, 8] neck disorders. standardised ROI method is arguably most important Such methods have provided convergent [9, 10] and for these studies where it has been speculated that divergent [11–15] evidence around insight into tissue difficulties identifying morphology of both the cervical composition, disease characterisation, response to injury, and lumbar musculature results in poorer repeatability and changes in somatic and nervous structures [6, 38]. With continued improvements in both the potentially due to biological, psychological, and socioen- uptake of, and imaging quality from, MRI technology, an vironmental stresses. Advances in MRI technology have agreed analysis plan utilising a common research raised the number of investigations quantifying skeletal measurement method for the identification of ROIs muscle composition (MFI) and structure (volume, cross- could result in meaningful comparisons with a target sectional area (CSA)), but not without equivocal results towards knowledge transfer and clinical translation of [9]. This variability in findings is likely the result of muscle imaging. Following on from the recent manuscript methodological differences across research groups, detailing a method for determining ROI in the lumbar including variables such as study design, participant spine [42], the purpose of this proposed method is to demographics (trauma vs. non-trauma; sex, sociocul- provide a standardised MRI procedure for measuring tural, age range), measurement techniques, and MR cervical spine muscle composition. The method also parameters used by investigators. serves to initiate and continue discussion on the analysis In order to better understand the influence of muscle of skeletal muscle composition amongst and between the composition and structure on cervical spine health, it is global clinical and scientific communities. imperative that clinical research communities explore and establish common methodologies in order to facili- Method tate standardisation and accurate comparison of data be- Challenges for producing a region of interest of cervical tween studies. Doing so should ultimately result in an muscles using MRI improved understanding of the aetiological features of A number of conventional MRI applications (T1, T2, muscle composition and facilitate an improved prognos- proton-density, Gradient Echo) are available and have tic, diagnostic, and theranostic landscape. been used to qualitatively and quantitatively measure the While data for age-related, degenerative changes of water and fat species of healthy and diseased soft-aqueous tissues (e.g. vertebrae, joints, discs, muscles) of the lum- skeletal muscle tissue [1, 3, 41, 43–49]. Technological bar and cervical spine have been published [16–27], advancements have also produced alternatives that can be studies assessing age-related alterations in paravertebral used to image muscle, such as dual acquisition methods, muscle morphology [19, 28, 29] remain unique to the where frequency is selectively excited to produce a water healthy lumbar spine. Such normative data, to our image [50] and a standard image of fat and water. This, knowledge, does not exist for the cervical spine. While however, produces a challenge when measuring a cross-sectional and longitudinal studies indicate a redundant and anatomically complex set of multi-layered positive relationship between MFI and traumatic neck (and small) muscles in the cervical spine. The challenge is pain (e.g. whiplash associated disorders) [1–3, 5, 6, 30], further compounded by the advent of higher field inconsistent associations are also reported [11–14]. Such scanners (e.g. 3–7 Tesla), where a uniform frequency inconsistencies have not improved our mechanistic un- difference between fat and water content may be difficult, derstanding of changes in muscle composition in both but certainly not impossible, to achieve. traumatic and non-traumatic neck pain. Future works Despite recent technological advances that have must collectively control for what might be considered permitted further insight into muscle composition, the normative age-related changes [19, 29, 31], degenerative mechanisms underlying muscle degeneration and their features of the vertebrae or discs [13, 26, 31–34], and influence on outcomes in neck disorders remain elusive. spinal curvature [35–37]. In addition, the vast majority of symptomatic and asymptomatic population-based studies examining A way forward through standardisation of methodology pathoanatomical features (e.g. the intervertebral disc, In order to facilitate widespread adoption of agreed and ligaments, and the skeletal vertebral column) of the time-efficient techniques for measuring cervical spine cervical spine have used a variety of conventional MRI muscle quality, a standardised, reliable, and replicable sequences [1, 2, 12, 13, 26, 30, 34, 51–55]. Despite the method is urgently required. While there is a general large repository of available works, the data derived from trend toward optimising automated methodologies that these imaging investigations have not revealed a consistent Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 3 of 14 structural lesion(s), or response to said lesion(s), that have The anatomical study we use and recommend for clarified the clinical presentation of traumatic or reference are those detailed in Au et al. [57]. They non-traumatic neck disorders. This has, in our have provided a comprehensive series of labelled axial opinion, created a clinical (and research) impasse MR images from one individual to serve as a that we believe is due partly to the heterogeneous reference atlas of the cervical spine musculature to methods across a number of high quality studies guide clinicians and researchers in the accurate investigating the usefulness of imaging for understanding identification of these muscles on MR imaging. We spinal pathology. Ultimately, the clinical value of imaging have further reinforced by cross-referencing with the findings of spinal pathology and/or muscle degeneration E-12 plastinates that have previously been used to will be realised if such findings predict important assist morphological studies [42, 58]. outcomes or help to identify patients likely to respond to specific interventions (e.g. spinal phenotypes). Anterior muscles Research efforts that focus on the consistent assessment Sternocleidomastoid (SCM) of spinal muscle quality with MRI may improve our The SCM arises from the manubrium and medial clavicle collective biological understanding of traumatic and inferiorly, and angles laterally and posteriorly towards its non-traumatic neck disorders and why some, but not superior attachments at the mastoid process and superior others, recover spontaneously. Accordingly, a robust nuchal line. This superficial muscle is readily identifiable and easily-replicated platform for acquiring, assessing, in cross-section. While the SCM has four portions, [59]as measuring, analysing, and interpreting imaging data they cross and blend, they are not separable in on muscle composition and morphology is needed. cross-section along their length on MRI. The muscle Currently a wide variety of methods are used to has an oval appearance inferiorly, and superiorly describe the composition and morphology of cervical forms a distinctive ‘comma’ shape (Fig. 1). spine muscles (see Table 1 for a non-exhaustive summary). This represents a key challenge for both Scalenus muscles producing consistent regions of interest of cervical Scalenus anterior arises from the scalene tubercle on the spine muscles and allowing comparison between first rib as a thin tendon antero-lateral to the lung and research studies. pleural cavities, and extends superiorly to attach to the anterior tubercles of the C4–6 (and frequently C3) transverse processes. At the level of the first rib the Anatomically defining the muscles of interest subclavian vein passes anterior to scalenus anterior, The muscles spanning the mid-to-lower cervical spine while the subclavian artery passes between scalenus that are typically examined include: multifidus, anterior and medius, visibly separating these two semispinalis cervicis, semispinalis capitis, splenius muscles. At this level scalenus anterior appears capitis, scalenes, levator scapulae, sternocleidomastoid, rounded in cross-section. Scalenus medius arises from and longus capitis and longus colli. We do not describe the first rib posterior to the groove for the subclavian muscles of the suboccipital region (rectus capitis artery and extends superiorly to attach to the posterior major and minor, and the superior and inferior transverse processes of C1–7. obliquus muscles [56]) as it is not possible to accurately measure a clinically useful ROI of the suboccipital Longus capitis muscles from the typically employed transverse images This muscle is largest at C1, and has a flattened appear- used for assessing cervical musculature. This is because ance immediately anterior to the lateral masses on each no suboccipital muscle has a long axis close to side of the midline. Inferiorly, it remains anterior to the perpendicular to the transverse plane, thus making anterior tubercles of the transverse processes, which al- measurement of useful cross-sectional ROI impractical. lows it to be differentiated from longus colli and the sca- Further, fan shaped muscles such as both rectus capitii lenus muscles, particularly scalenus anterior (Fig. 2)[60]. muscles require special consideration in order to validate useful measures, given a single cross-sectional measure- Longus colli ment along the length of either muscle would pose diffi- Longus colli is recognised by its location in the groove culty for determining whole muscle volume. Future formed between the vertebral bodies and transverse pro- work should include developing imaging protocols for cesses of the vertebrae, extending between C1 and T2/3. the suboccipital muscles as they require more nuanced While longus colli is described as having superior, verti- imaging methods and measures with careful consider- cal, and inferior oblique portions, these are based on at- ation around the highest resolution possible within a tachment sites and are not discernible in cross-section reasonable scan time. [61]. The muscle first becomes visible at C2, emerging Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 4 of 14 Table 1 A non-systematic summary of methods across investigations describing cervical spine muscle analysis using magnetic resonance imaging (MRI) Citation Reliability MRI Slice Selection Muscles of Interest ROI Selection Fat Detection Measure Sequence Elliott et al., Inter-rater T1 Axial images aligned parallel to C2–3 disc; MF Manual Quantitative Pixel Fat Infiltration 2006 [1] (0.94) Measured at single slice per level C3-C7; SSCerv Intensity Elliott et al., Intra-rater most cephalad slice of each vertebral SSCap %MFI = (muscle signal)/ 2008 [7] (0.94) body selected SpCap (fat signal)*100 Elliott et al., UT 2009 [78] Elliott et al., 2011 [2] Fernandez Inter-rater T1 Axial images aligned parallel to C2–3 disc; SSCap Manual N/A CSA De Las Penas (0.80–0.98) measured at single slice per level; most SpCap et al., 2007 [79] cephalad slice of each vertebral body selected Elliott et al., Intra-rater T1 Axial images aligned parallel to C2–3 disc; MF Manual N/A CSA 2007 [47] (0.84–0.99) measured at single slice per level C3-C7; SSCerv Inter-rater most cephalad slice of each vertebral SSCap (0.89–0.96) body selected SpCap UT Okada et al., Intra- rater T2 Measurements from a single axial slice MF Manual N/A CSA 2011 [80] (0.90) aligned parallel to each IVD C3–4, C4–5, SSCerv Matsumoto Inter-rater and C5–6 SSCap et al., 2012 [12] (0.844) SpCap Ulbrich et al., Inter-rater STIR Axial images aligned perpendicular to Deep Extensors Manual N/A CSA 2012 [14] (0.79–0.98) the vertebral body in the middle of a All Extensors 20-slice slab. 2 or 3 overlapping slabs SCM used; measurements from single slice per vertebral level C2, C4, and C5 Elliott et al., Inter-rater T1 vs. Dixon Axial images aligned perpendicular to MF Manual Quantitative Pixel Fat Infiltration 2013 [41] for fat- the spinal cord at the C2-C3 IVD; Intensity vs. Fat water sequence measurement from single slice per (F)-Water (W) % MFI = (fat signal)/(fat (0.83–0.99) vertebral level C3-C7 signal + water signal) *100 Elliott et al., From Elliott T1 Axial images aligned parallel to MF Manual Quantitative Pixel Fat Infiltration 2014 [8] et al., 2007 [47] C2–3 IVD; measurements from single SSCerv Intensity for CSA slice crossing IVDs C2-C3 and C5-C6 SSCap %MFI = (fat signal)/(fat Intra- rater SpCap signal + water signal)*100 (0.84–0.99) LCap/LCol Inter-rater SCM CSA (0.89–0.96) Elliott et al., Not reported Dixon Measurements from single slice per MF Manual Fat-Water Fat Infiltration 2015 [3] vertebral level C3-C7; alignment and slice selection not reported %MFI = (fat signal)/(fat signal + water signal)*100 Abbott et al., Intra-rater (0.98) Dixon Measurements averaged over 5 slices MF + SSCerv Manual with Fat-Water Fat Infiltration 2015 [6] Inter-rater (0.93) for each vertebral level C3-C7; Slab (combined) automatic alignment not reported. quartile measure %MFI = (fat signal)/(fat signal + water signal)*100 Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 5 of 14 Table 1 A non-systematic summary of methods across investigations describing cervical spine muscle analysis using magnetic resonance imaging (MRI) (Continued) Citation Reliability MRI Slice Selection Muscles of Interest ROI Selection Fat Detection Measure Sequence Karlsson et al., For muscle fat Dixon Axial images aligned parallel to vertebral MF Manual Fat-Water Fat Infiltration 2016 [5] Intra-rater segments; measurements averaged over (0.81–0.93) 5 slices for each vertebral level C4-C7 %MFI = (fat signal)/(fat Inter-rater signal + water signal)*100 (0.82–0.97) CSA Au et al., 2016 Not reported T1 Axial images aligned parallel to C2–3 IC, IS, LS, LoCap, Manual N/A N/A [57] intervertebral disc; 3D reconstruction LoC, LCap, LCol, MF, LoCap, LoCerv, SSCap, SSCerv, SCM, UT Fortin et al., From [81]: T2 3D multiplanar reconstruction to align MF Manual ROI with Gray-scale threshold Total CSA 2017 [27] Intra-rater images perpendicular to muscle mass; SSCerv semi-automatic technique to calculate Fortin et al., measurements from a single slice per SSCap muscle/fat CSA of fat within Functional CSA (FCSA) 2018 [81] (0.83–0.99) IVD C2–3 through C6–7 SpCap thresholding total muscle CSA; (= fat free area) Inter-rater technique gray-scale range Fatty Infiltration = FCSA/ determined for Total CSA (0.38–0.98) each slice individually Inoue et al., Intra-rater T1 Measured from single slice per level; most MF Manual Lean muscle CSA: ROI Fatty Infiltration = (Fat 2012 [82] (0.85) T2-fat caudal slice of C3 and most cephalad slice drawn on T1-W images CSA)/(Total muscle CSA) suppression of each vertebral body C4-C7 selected; slab not including fat alignment not reported Total muscle CSA: ROI drawn on fat suppression T2-W image including fat Fat CSA = Total CSA – Lean Muscle CSA Mitsutake Intra-rater T1 Measured from single, most cephalad slice MF Manual Quantitative Pixel MFI index = Muscle et al., 2016 0.85–0.94 at level of injury (C4, C5, or C6) Intensity signal/Fat signal [83] Inter-rater (0.84–0.89) Abbott et al., Intra-rater Dixon Axial images aligned parallel to each MF Manual Qualitative grading Fat Infiltration 2017 [67] (0.77–0.88) IVD; measured from 5 slices across each (0, 1, 2) for each 8 MFI Score: Inter-rater vertebral level C4-C7 regions within 0 = no or marginal fat (0.67–0.82) visualized ROI on 1 = light fat fat image 2 = distinct fat Sum of scores Total # of 2’s Choi et al., Inter-rater T1 Axial images aligned parallel to the inferior Flexor Group: Manual N/A CSA 2016 [84] (0.82) end plate of each vertebral body from C4–5 LCap + LCol Normalized Extensor to C7-T1; measured from single slice per Extensor Group: CSA = (Extensor Muscle vertebral level MF + SSCerv CSA)/(Vertebral body CSA) *100 Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 6 of 14 Table 1 A non-systematic summary of methods across investigations describing cervical spine muscle analysis using magnetic resonance imaging (MRI) (Continued) Citation Reliability MRI Slice Selection Muscles of Interest ROI Selection Fat Detection Measure Sequence Cagnie et al., Inter-rater T1 Measured from a single slice aligned LCap Manual Quantitative Pixel Muscle/Fat Index = Muscle 2009 [85] (0.91) parallel IVD at C4-C5 LCol Intensity signal/Fat signal UT LS SpCap SSCerv MF Uthaikup et al., Intra-rater T1 Axial images aligned parallel to the MF Manual Quantitative Pixel Intensity Fat Infiltration 2017 [86] (0.75–0.96) C2–3 IVD; measured from a single SSCap Inter-rater slice at each vertebral level C2-C3 SpCap MFI = Muscle signal/ (0.84–0.99) LCap Fat signal LCol SCM Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 7 of 14 Fig. 1 Axial E12 plastinated section (a) with schematic illustration (b) Fig. 2 Axial E12 plastinated section (a) with schematic illustration and in-phase magnetic resonance image (c) at approximately C2/3 (b) and in-phase magnetic resonance image (c) at approximately identifying musculature at this vertebral level. 1. Longus colli; 2. Longus C5/6 identifying musculature at this vertebral level. Dashed red capitis; 3. Intertransversarii; 4. Levator scapulae; 5. Sternocleidomastoid; (b) and white (c) line indicates an anatomical plane which can 6. Longissimus capitis; 7. Splenius cervicis; 8. Inferior obliquus; 9. Rectus be used as a reference point for identifying some anterior muscles. capitis posterior major; 10. Semispinalis capitis; 11. Splenius capitis; Dashed white line in (c) indicates likely border between multifidus and 12. Trapezius semispinalis cervicis. 1. Sternocleidomastoid; 2. Longus colli; 3. Longus capitis; 4. Scalenus anterior; 5. Scalenus medius; 6. Splenius cervicis; 7. Multifidus / semispinalis cervicis; 8. Semispinalis capitis; 9. Splenius capitis; medial to longus capitis and initially with a more 10. Levator scapulae; 11. Trapezius rounded appearance. Inferior to C7 the muscle thins and moves towards the midline, before attaching to the an- terolateral vertebral bodies. Fascial borders between muscles are only seen in slices between transverse pro- longus colli and the intertransversarii muscles may not cesses (Fig. 1). be readily apparent between any of the cervical levels on MRI. This should not, however, present difficulties as Posterior muscles long as the bony transverse processes are well visualised. Multifidus and rotatores Longus colli remains immediately anterior and medial to Deep against the vertebra, these architecturally complex the bony transverse processes. The intertransversarii muscles fill the space between the spinous and transverse Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 8 of 14 processes. Multifidus is present along the length of the between these two muscles [62], it can be difficult to spine below C2, forming the deepest layer (Figs. 2, 3). identify them as separate entities on both E12s and MRI. Rotatores can be considered together with multifidus in this deep muscle layer, as these muscles are small and do Semispinalis cervicis not form a distinct layer able to be identified in cross- Semispinalis cervicis extends between the spinous section. Together with semispinalis cervicis, multifidus sits processes of C2–5 and the transverse processes of T1- in the paravertebral gutter between the spinous and trans- T5 [63] (Figs. 1, 2, 3). It overlies multifidus along with verse processes. Because of the intimate relationship other cervical-attaching erector spinae (longissimus cervicis, iliocostalis cervicis). The semispinalis cervicis and erector spinae muscles are difficult, if not impos- sible, to adequately distinguish in cross-section. The close approximation, similar alignment and attachments of multifidus, semispinalis and erector spinae fascicles are such that a distinct layer will not always be clear on MRI. In this situation it is reasonable to consider these muscles together as a single group (seen [64] and [6]). Semispinalis capitis Semispinalis capitis is a major muscle of the cervical spine, overlying semispinalis cervicis and forming a large and distinct muscle layer. While semispinalis capitis spans between the occiput and T6–7[63], in cross- section this layer is most apparent between the occiput and C6/7. Below this level this muscle layer becomes less distinct as semispinalis thins and becomes tendinous to- wards the thoracic transverse processes (Figs. 1, 2). Erector spinae Longissimus cervicis extends between the thoracic transverse processes of T1–4 and the C2–6transverse processes, while iliocostalis cervicis passes between the angles of ribs 3–4 and the transverse processes of C4–6 [63]. As noted, erector spinae muscles attaching to the cervical spine are unlikely to be differentiated from semi- spinalis cervicis. Longissimus capitis is more distinct, ex- tending between the mastoid process and the transverse processes of approximately C4-T4 (Fig. 1)[63]. Splenius capitis and cervicis Splenius capitis and cervicis form a single layer and overlie semispinalis capitis. Splenius capitis spans be- tween the C7-T4 spinous processes and the mastoid process / occiput, while splenius cervicis spans between the T3–6 spinous processes and the transverse processes of C1–3[63]. In cross-section, splenius capitis forms a Fig. 3 Axial E12 plastinated section (a) with schematic illustration (b) distinct layer between trapezius and semispinalis capitis. and in-phase magnetic resonance image (c) at approximately C7/T1 Splenius cervicis can be identified between C2–6 on the identifying musculature at this vertebral level. Red box indicates antero-lateral edge of this layer (Figs. 1, 2, 3), as it boundary for Fig. 4. 1. Sternocleidomastoid; 2. Scalenus anterior; 3. Longus colli; 4. Scalenus medius; 5. Iliocostalis cervicis; 6. Multifidus / diverges from splenius capitis towards its cervical attach- semispinalis cervicis; 7. Serratus posterior superior; 8. Splenius capitis ments. Below the level of approximately C5 splenius / cervicis; 9. Levator scapulae; 10. Serratus anterior; 11. cervicis is unlikely to be visibly separate from splenius Rhomboid minor capitis in cross-section. Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 9 of 14 Levator scapulae Levator scapulae have a presence throughout the cervical spine, and its presence is worth noting as one of the larger and more distinctive muscles in the region. It passes from the upper aspect of the medial scapula to the transverse processes of C1–4[63]. In cross-section levator scapulae is well-defined at lower levels, sitting anterior to trapezius and lateral to splenius (Figs. 2, 3). Superiorly, levator scapulae extends towards the trans- verse processes of C1–4 in close relation to the scalenus and longus capitis muscles (Fig. 1). Results Our method provides anatomical reference between MRI imaging and E12 plastinates (derived from cadavers) to advance ROI identification and definition to improve standardised measurement of musculature traversing the cervical spine. The E12 plastinates provide a unique opportunity to detail specific tissues that may be MR invisible, [65] leading to natural disagreement across studies where fat-water separation is a target. To follow, we also include suggestions on operational Fig. 4 Enlarged region of posterior cervical muscles from Fig. 3 (c), characteristics for acquiring MR images. highlighting deep boundary of region of interest (white solid line along lamina). Common mistakes for determining this region of Defining the regions of interest from MRI interest for the transversospinal (TSP) muscles include the boundary Similar to that reported for the lumbar spine, [42]a of multifidus and semispinalis cervicis (white dashed line) or through standard scout image from the sagittal localiser or the fatty infiltrate in multifidus (red dashed line) conventional T2-weighted scan can be used to cross- reference and discern cervical level from axial MR. Users will also find it useful to scroll between the adjacent point for identifying the anterior aspect of all anterior axial slices to accurately landmark anatomical structures muscles apart from the sternocleidomastoid (Fig. 2). when producing ROIs. The method is applicable to studies examining paravertebral ROIs for single (cross- Sternocleidomastoid sectional) or multiple (volumetric) slices. Previous work This definition can be applied along the full extent of from the lumbar spine suggests a randomised approach sternocleidomastoid, from the occiput to approximately for starting with either the left or right side, and/or sep- T2/3. The anatomical boundaries of sternocleidomastoid arate muscles can influence repeatability when creating are straight forward, and tracing should present few ROIs [38, 66]. The same randomised approach is challenges. Some care is needed to trace along the full suggested for the cervical spine. occipital extent at higher levels (Fig. 1). Definitions for ROI measures from MRI for the multi- fidus, semispinalis cervicis, semispinalis capitis, longissi- Scalenus muscles mus capitis, splenius capitis and cervicis, levator This definition is best applied at the C6-T2 levels. The scapulae, longus colli, longus capitis, scalenus and scalenus muscles are best identified at their inferior sternocleidomastoid are included, describing the ana- extent arising from the first rib. Superiorly scalenus tomical borders (cross referenced to Figs. 1, 2, 3). ROI anterior and scalenus medius converge, and may be definitions are detailed with particular reference to cer- difficult to differentiate above the level of C6 on MRI. vical levels C2/3, C5/6, and C7/T1. Technical notes are Differentiation is aided by the angle each muscle ap- also provided where identifying the guided ROI on MRI proaches the cervical transverse processes, as each may be difficult. muscle follows a straight course. Sequentially from anterior to posterior: longus capitis is seen anterior to Anterior muscles the anterior tubercles, scalenus anterior angles to attach It is worth noting that an anatomical plane that passes to the anterior tubercles slightly more laterally, scalenus laterally and posteriorly in an arc from the anterior medius angles between the anterior and posterior tuber- aspect of the vertebral body presents a reliable reference cles, scalenus posterior (if present) angles towards the Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 10 of 14 posterior tubercles, and (above C4) levator scapulae also Semispinalis capitis angles to attach to the posterior tubercles (Fig. 2). This definition is best applied at the occiput-C6 levels. The muscle forms a distinct anatomical layer and can be traced consistent with the anatomy described. As semi- Longus capitis spinalis capitis is clearest at higher levels, difficulties This definition is best applied at the C1–5 levels. Longus identifying this muscle at lower cervical levels would capitis is largest and most distinct superiorly, just below benefit from reviewing and toggling between multiple where it attaches to the basi-occiput. As such, the slices from superior to inferior. As the E12 slices high- muscle is best tracked inferiorly from this point. At its light, it may not be realistic to identify this muscle below superior extent longus capitis has a rounded appearance, approximately C7. which flattens and thins out over the lateral masses of C1. By the level of C2/3 longus capitis is a relatively thin slip immediately anterior to the anterior tubercles of Longissimus capitis transverse processes C3–6. As for the scalenus muscles, This definition is best applied at the C1–4 levels. Longis- identification is aided by identifying the transverse pro- simus capitis is most easily identified as a rounded cesses (in particular the anterior tubercles) and muscle at its superior extent, just below the mastoid remaining posterior to the prevertebral fascia (Fig. 2). process. Towards C4 the muscle flattens, and below ap- proximately C4 it becomes difficult to distinguish from Longus colli other muscles. This definition is best applied at C2-T1 levels. As noted anatomically, longus colli sits in the groove between the vertebral bodies and transverse processes of the verte- Splenius capitis and cervicis brae. Thus, these bony landmarks must be well visua- This definition is best applied at the C1-T3 levels. Sple- lised to accurately identify the muscle. As described for nius capitis is identifiable as a distinct layer located be- multifidus, the ROI should closely follow the bony verte- tween trapezius / sternocleidomastoid and the brae to include fat adjacent to the bone. If the anterior semispinalis capitis. Care is needed around the level of aspect of the transverse processes are not visible, or the mastoid process not to confuse the superior extent slices above and below are not reviewed to clarify the of splenius capitis with sternocleidomastoid or longissi- position of bony landmarks, a ROI for longus colli may mus capitis, which share attachment to the mastoid be inaccurate. process. Just below the mastoid process at the level of C1/2 the muscles appear closely layered from superficial to deep: sternocleidomastoid, splenius capitis (both an Posterior muscles elongated comma shape), and longissimus capitis Multifidus and semispinalis Cervicis (rounded in appearance). Below this level the muscles This definition is best applied at the caudal portion of diverge. Ideally, splenius cervicis will be able to be dis- the C4 vertebral body through the entire T1 vertebral tinguished from splenius capitis at the levels of C2–6 body. With current technology it is generally not pos- (Figs. 1, 2). However, this may not be realistic with sible to consistently delineate between the cervical por- current MRI technology. In this situation, it is reason- tions of the semispinalis cervicis and multifidus on the able to include splenius capitis and cervicis together as a axial slices. While measuring the two muscles independ- single ROI. ently is recommended, they can be combined to form one measure (Figs. 2, 3). As evidenced from the lumbar spine [42], the same approach of approximating the spinous process or lamina is recommended and should Levator scapulae be included within the ROI defining cervical multifidus This definition is best applied at the C2- T1 levels. (Fig. 4). A challenge for both novice and expert clini- While not part of the intrinsic cervical spine muscula- cians remains what to do when creating ROIs between ture, cross-sectional views highlight the presence and the spinous processes. Whether referencing the lumbar size of levator scapulae throughout the cervical spine. [42] or cervical spine, fat commonly overlies the inter- This muscle is largest inferiorly above where it arises spinous space, remains defined, and should be included from the upper part of the medial scapula border, and as when generating ROIs on these slices. Finally, when the such is best tracked superiorly from this point. Care is interspinous ligaments are clearly distinct with a slightly needed to distinguish levator scapulae from serratus irregular and darkened edge, their lateral contour can be anterior as they converge on the scapula (Fig. 3). Atten- followed rather than the spinous process in defining the tion to slices above and below the level of interest will medial border [42]. help resolve their borders. Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 11 of 14 MR imaging - operational parameters [3, 5, 6, 27, 41]. Such an approach has revealed not The type, quality, and output of images acquired from only improved inter- and intra-rater reliability when MR scans are highly influenced by many factors includ- following the spinous process and/or lamina in the ing, but not limited to, user-prescribed parameters. Simi- cervical spine, but also the ability to discriminate be- lar to our previous paper covering the lumbar spine, [42] tween clinical groups [6]. This improved repeatability we endorse consistency in the adoption of MR imaging for defining MF over ES in the lumbar spine has also parameters to facilitate standardised operational proce- been demonstrated [38]. dures that allow intra-study/−institutional comparison and future pooling of results for meta-analyses. The parameters listed here are based on those widely Measures of muscle size and fat utilised in literature (refer to Table 1), and are adapted Measures of muscle size are frequently reported in MRI from those published in a previous paper on ROI for and other imaging-based studies (e.g. ultrasound). In lumbar spine muscles [42]. The parenthetical values both the lumbar and cervical regions, methods employ- given with each parameter are not definitive or unique ing a single cross-sectional MR slice are time efficient to a cervical spine study; rather they are displayed as an for determining muscle size and fat proportion within example of the consistent reporting style we propose. At an ROI. However, a CSA measure from a single-slice a minimum, we believe the following information should should not be taken to constitute a whole muscle size or be reported in all submitted manuscripts: Field strength fat measure [15, 68]. Accordingly, volumetric measures, (e.g. 3 Tesla); sequence type (e.g. 2-point DIXON (3D may be more appropriate [15, 69, 70]. We therefore rec- fast-field echo T1) whole body); repetition time (e.g. TR ommend a multi-slice approach that derives muscle size 4.2 ms); echo time (e.g. TE 1.2 and 3.1 ms); flip angle (e. and fat content based on a three-dimensional volume g. 5°); field of view (e.g. FOV 560 × 352 mm); acquired across the levels of interest. In going forward, such voxel dimensions (e.g. 2.0 × 2.0 × 4.0 mm); reconstructed measures should be accurately categorised as a 3- voxel dimensions (e.g. 1.0 × 1.0 × 2.0 mm); bandwidth (e.g. dimensional volume of the entire muscle as 3D acquisi- 240 Hz/Px), acquisition time (e.g. TA 5 min 22 s) and slice tion methods with MRI have evolved and are not as thickness (e.g. 4.0 mm). Additionally, the description sensitive to the radio frequency slice profile as is 2D should include axial slice alignment (e.g. aligned parallel imaging [15]. to C2–3 intervertebral disc), slice selection (e.g. measure- It is of course acknowledged that acquiring such data ments taken at most cephalad slice per vertebral level), with both semi-automated or automated programmes and subject body position including any support materials for both the lumbar [42] and cervical spines is time- that may influence cervical spine posture/curvature (e.g. consuming. However, with the evolution of higher- subjects positioned supine with arms at sides and 2 in. resolution imaging techniques a more time-efficient foam cushion under head). capture of cervical muscle volumes from a single verte- bral level may correspond to a representative marker of Discussion MFI across the entire cervical column. While this has A foundational edict for defining lumbar paravertebral been demonstrated in the healthy lumbar spine [29] ROI’s from MRI studies has previously been published where the fat content at L4 best represents that of the [42]. Here, we expand the previous methods [57] for the entire lumbar region, future research should continue to cervical pre- and para-vertebral muscles using a number systematically include the entire cervical spine in healthy of MRI and E12 sheet plastinate illustrations of vertebral and symptomatic cohorts to build a stronger body of morphology with the aim of standardising muscle ROI evidence regarding age-aggregated cervical paravertebral definitions. The E12 plastinates provide a unique oppor- muscle composition. tunity to detail specific tissues that may be MR invisible, Another issue with longitudinal designs, where muscle [65] leading to natural disagreement across studies measures are produced over time, remains a general lack where fat-water separation is a target. Also unique to of reporting on how the MRI slices are aligned in plane. this work is the included suggestions on operational A failure to do so could potentially result in registration characteristics for acquiring MR images. discrepancies depending how the angle through each Similar to the proposed approach in the lumbar spine, muscle was performed. Using some standard anatomical [42] we consider that if fat is occupying space deep to reference (e.g. vertebral bone) that is not expected to ap- the epimysial sheath and close to the spinous processes, preciably change over time could control for this. Errors laminae, zygapophyseal joints, it has a potential bio- of this type can be further minimised by reporting mechanical consequence on muscle function, [6, 67] and muscle volume over the full length of the muscle (from should be included in the ROI (Fig. 4). We base this de- origin to insertion), as suggested above, rather than a cision in part on previous work in the cervical spine single-slice CSA. Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 12 of 14 Measures of muscle fat with MRI expanded, and refined and ultimately result in an estab- The demonstration of neck muscle fatty infiltrates on lished common methodology towards facilitating con- T1- weighted imaging in acute [2, 3] and chronic trau- sistent and accurate definitions of lumbar, cervical, and matic neck pain [1, 5, 8, 30] has been reported in cross- upper/lower limb muscle ROIs on axial imaging, sectional and longitudinal fashion and across three particularly MRI. countries (Australia, [2] Sweden, [5] and the United Abbreviations States [3, 6]). Such findings are not present to the same CSA: Cross-sectional area; FCSA: Functional cross-sectional area; magnitude for those with chronic idiopathic neck pain FSPGR: Fast-spoiled gradient echo; IC: Iliocostalis cervicis; IP: In-phase (water); IS: Interspinalis cervicis; IVD: Intervertebral disc; LCap: Longus [7] and it has been postulated that these muscle changes capitis; LCol: Longus colli; LoCap: Longissimus capitis; LoCerv: Longissimus represent one neurophysiologic basis for the transition cervicis; LS: Levator scapulae; MF: Multifidus; MFI: Muscle fat infiltration; to chronic pain in this population [71]. A variety of OP: Opposed-phase (Fat); SCap: Spinalis capitis; SCerv: Spinalis cervicis; SCM: Sternocleidomastoid; SI: Signal intensity; SpCap: Splenius capitis; newer and more rapid high resolution MRI techniques SSCap: Semispinalis capitis; SSCerv: Semispinalis cervicis; UT: Upper trapezius (3D Fat/Water Separation and Proton-Density Fat Frac- tion, Fat suppression) [65, 72–77] and analyses (FCSA/ Availability of data and materials CSA, Fat Signal Fraction, MFI %) could help better visu- All the data supporting the findings are contained within the manuscript. alise and quantify physiologic changes at the level of the Authors’ contributions muscle cell or other disease processes when compared JME and RJC conceived the study, while RJC, JC, EK, RA, and JME each made to other conventional clinical imaging sequences (e.g. substantial contributions to its design. RJC, JC, EK, RA, and JME have been T1- and T2-weighted). However, such variety across involved in drafting the manuscript or revising it critically for important intellectual content and each given final approval of the version to be methods and techniques also complicates comparison published. RJC, JC, EK, RA, and JME agree to be accountable for all aspects of among studies. Accordingly, we call for all authors to the work in ensuring that questions related to the accuracy or integrity of clearly detail their fat infiltration measurements to any part of the work are appropriately investigated and resolved.All authors read and approved the final manuscript. ensure that future pooling of data efforts is possible. Fur- ther, with the number of proprietary semi-automated or Ethics approval and consent to participate automated methods appearing in the literature, and of No ethics approval was required to undertake this descriptive methodological study. The preparation of the E-12 plastinated sections took which descriptions are limited due to commercial sensi- place in accordance with the New Zealand Human Tissue Act (2008) and tivity, we contend it will be helpful for authors to in- University of Otago institutional guidelines. clude enough technical detail for comparisons to the fundamental literature to be made. Consent for publication Approval to use images of the E-plastinated sections was granted by Depart- ment of Anatomy, University of Otago. All MRIs were derived from the same Participant positioning informed and consenting adult subject. It is our recommendation that participants should lie su- Competing interests pine inside the magnet with a foam pad under their Authors RJC, JC, EK, and RA have no disclosures to declare. In unrelated knees and foam padding placed on the right and left of activities, JME is principal investigator on NIH grant [HD079076-01A1; 09/ the head to minimise head movement. A neutral pos- 2014–05/2019]. ition, visually determined by ensuring that a horizontal position of the forehead to the chin is parallel to the Publisher’sNote MRI table, is also recommended. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Conclusion Author details We follow on from, and have expanded, an original Faculty of Health Sciences, The University of Sydney, Northern Sydney Local Health District, St Leonards, Australia 75 East Street Lidcombe NSW, Sydney paper of manually defining ROIs of lumbar spine mus- 2141, Australia. Department of Physical Therapy and Human Movement culature [42] to now include the cervical muscles. While Sciences, Feinberg School of Medicine, Northwestern University, Chicago, the method aims to permit accurate and reliable com- USA. Honorary Fellow School of Health and Rehabilitation Sciences, The University of Queensland, St. Lucia, Australia. Centre for Early Learning in parison of cervical muscle quality between studies in Medicine, Otago Medical School, University of Otago, Dunedin, New Zealand. (and beyond) this field, we further suggest journals adopt School of Physiotherapy, University of Otago, Dunedin, New Zealand. a more robust reporting of imaging parameters used to Faculty of Health Sciences, Curtin University, Perth, Australia. assist consistency and allow accurate comparison be- Received: 14 February 2018 Accepted: 4 May 2018 tween studies. It is imperative to note that we are cognisant the application methods are not definitive end-points on References 1. Elliott J, Jull G, Noteboom JT, Darnell R, Galloway G, Gibbon WW. Fatty ‘how to’ and that there is potential for much repetition infiltration in the cervical extensor muscles in persistent whiplash-associated across body regions. Rather, we hope that with time, and disorders: a magnetic resonance imaging analysis. Spine (Phila Pa 1976). new research findings, these methods will be modified, 2006;31(22):E847–55. Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 13 of 14 2. Elliott J, Pedler A, Kenardy J, Galloway G, Jull G, Sterling M. The temporal 22. Hancock MJ, Kjaer P, Kent P, Jensen RK, Jensen T. Is the number of different development of fatty infiltrates in the neck muscles following whiplash MRI findings more strongly associated with low back painthan single MRI injury: an association with pain and posttraumatic stress. PLoS One. 2011; findings? Spine. 2017;42(17):1283–8. 6(6):e21194. 23. Panagopoulos J, Hush J, Steffens D, Hancock M. Do MRI findings change 3. Elliott JM, Courtney DM, Rademaker A, Pinto D, Sterling MM, Parrish T. The over a period of up to one year in patients with low back pain and/or rapid and progressive degeneration of the cervical multifidus in whiplash: a sciatica? A Systematic Review. Spine. 2017;42(7):504–12. MRI study of fatty infiltration. Spine (Phila Pa 1976). 2015;40(12):E694–700. 24. Steffens D, Hancock MJ, Maher CG, Williams C, Jensen TS, Latimer J. Does 4. Elliott JM, Pedler AR, Theodoros D, Jull GA. Magnetic resonance imaging magnetic resonance imaging predict future low back pain? A systematic changes in the size and shape of the oropharynx following acute whiplash review. Eur J Pain. 2014;18(6):755–65. injury. J Orthop Sports Phys Ther. 2012;42(11):912–8. 25. Wan Q, Lin C, Li X, Zeng W, Ma C. MRI assessment of paraspinal muscles in 5. Karlsson A, Dahlqvist Leinhard O, West J, Romu T, Aslund U, Smedby O, patients with acute and chronic unilateral low back pain. Br J Radiol. 2015; Zsigmond P, Peolsson A. An investigation of fat infiltration of the multifidus 88(1053):20140546. muscle in patients with severe neck symptoms associated with chronic 26. Nakashima H, Yukawa Y, Suda K, Yamagata M, Ueta T, Kato F. Abnormal whiplash associated disorder. J Orthop Sports Phys Ther. 2016;46(10):886–93. findings on magnetic resonance images of the cervical spines in 1211 6. Abbott R, Pedler A, Sterling M, Hides J, Murphey T, Hoggarth M, Elliott J. asymptomatic subjects. Spine (Phila Pa 1976). 2015;40(6):392–8. The geography of fatty infiltrates within the cervical multifidus and 27. Fortin M, Dobrescu O, Courtemanche M, Sparrey CJ, Santaguida C, Fehlings semispinalis Cervicis in individuals with chronic whiplash-associated MG, Weber M. Association between paraspinal muscle morphology, clinical disorders. J Orthop Sports Phys Ther. 2015;45(4):281–8. symptoms and functional status in patients with degenerative cervical 7. Elliott J, Sterling M, Noteboom JT, Darnell R, Galloway G, Jull G. Fatty myelopathy. Spine. 2017;42(4):232–9. infiltrate in the cervical extensor muscles is not a feature of chronic, 28. Amabile C, Moal B, Chtara OA, Pillet H, Raya JG, Iannessi A, Skalli W, Lafage V, insidious-onset neck pain. Clin Radiol. 2008;63(6):681–7. Bronsard N: Estimation of spinopelvic muscles' volumes in young asymptomatic 8. Elliott JM, Pedler AR, Jull GA, Van Wyk L, Galloway GG, O’Leary S. Differential subjects: a quantitative analysis. Surgical and radiologic anatomy: SRA. 2016;39(4): changes in muscle composition exist in traumatic and non-traumatic neck 393–403. pain. Spine (Phila Pa 1976). 2014;39(1):39–47. 29. Crawford R, Filli L, Elliott J, Nanz D, Fischer M, Marcon M, Ulbrich E. Age- 9. De Pauw R, Coppieters I, Kregel J, De Meulemeester K, Danneels L, Cagnie and level-dependence of fatty infiltration in lumbar paravertebral muscles of B. Does muscle morphology change in chronic neck pain patients? - a healthy volunteers. Am J Neuroradiol. 2016;37(4):742–8. systematic review. Man Ther. 2016;22:42–9. 30. Elliott JM, O'Leary S, Sterling M, Hendrikz J, Pedler A, Jull G. Magnetic 10. Thakar S, Mohan D, Furtado SV, Sai Kiran NA, Dadlani R, Aryan S, Rao AS, resonance imaging findings of fatty infiltrate in the cervical flexors in Hegde AS. Paraspinal muscle morphometry in cervical spondylotic chronic whiplash. Spine (Phila Pa 1976). 2010;35(9):948–54. myelopathy and its implications in clinicoradiological outcomes following 31. Bhadresha A, Lawrence OJ, McCarthy MJ. A comparison of magnetic central corpectomy: clinical article. J Neurosurg Spine. 2014;21(2):223–30. resonance imaging muscle fat content in the lumbar Paraspinal muscles with 11. Anderson SE, Boesch C, Zimmermann H, Busato A, Hodler J, Bingisser R, patient-reported outcome measures in patients with lumbar degenerative disk Ulbrich EJ, Nidecker A, Buitrago-Tellez CH, Bonel HM, et al. Are there cervical disease and focal disk prolapse. Glob Spine J. 2016;6(4):401–10. spine findings at MR imaging that are specific to acute symptomatic 32. Teichtahl AJ, Urquhart DM, Wang Y, Wluka AE, Wijethilake P, O'Sullivan R, whiplash injury? A prospective controlled study with four experienced Cicuttini FM. Fat infiltration of paraspinal muscles is associated with low blinded readers. Radiology. 2012;262(2):567–75. back pain, disability, and structural abnormalities in community-based 12. Matsumoto M, Ichihara D, Okada E, Chiba K, Toyama Y, Fujiwara H, adults. Spine J. 2015;15(7):1593–601. Momoshima S, Nishiwaki Y, Takahata T. Cross-sectional area of the posterior 33. Kalichman L, Hodges P, Li L, Guermazi A, Hunter DJ. Changes in paraspinal extensor muscles of the cervical spine in whiplash injury patients versus muscles and their association with low back pain and spinal degeneration: healthy volunteers - 10year follow-up MR study. Injury. 2012;43(6):912–6. CT study. Eur Spine J. 2010;19(7):1136–44. 13. Matsumoto M, Okada E, Ichihara D, Chiba K, Toyama Y, Fujiwara H, 34. Matsumoto M, Ichihara D, Okada E, Toyama Y, Fujiwara H, Momoshima S, Momoshima S, Nishiwaki Y, Hashimoto T, Inoue T, et al. Prospective ten-year Nishiwaki Y, Takahata T. Modic changes of the cervical spine in patients with follow-up study comparing patients with whiplash-associated disorders and whiplash injury: a prospective 11-year follow-up study. Injury. 2013;44(6):819–24. asymptomatic subjects using magnetic resonance imaging. Spine (Phila Pa 35. Meakin JR, Fulford J, Seymour R, Welsman JR, Knapp KM. The relationship 1976). 2010;35(18):1684–90. between sagittal curvature and extensor muscle volume in the lumbar 14. Ulbrich EJ, Aeberhard R, Wetli S, Busato A, Boesch C, Zimmermann H, spine. J Anat. 2013;222(6):608–14. Hodler J, Anderson SE, Sturzenegger M. Cervical muscle area measurements 36. Pezolato A, de Vasconcelos EE, Defino HL, Nogueira-Barbosa MH. Fat in whiplash patients: acute, 3, and 6 months of follow-up. J Magn Reson infiltration in the lumbar multifidus and erector spinae muscles in subjects Imaging. 2012;36(6):1413–20. with sway-back posture. Eur Spine J. 2012;21(11):2158–64. 15. Elliott JM, Kerry R, Flynn T, Parrish T. Content not quantity is a better 37. Johansson MP, Baann Liane MS, Bendix T, Kasch H, Kongsted A. Does measure of muscle degeneration in whiplash. Man Ther. 2013;18(6):578–82. cervical kyphosis relate to symptoms following whiplash injury. Man Ther. 16. Brinjikji W, Luetmer PH, Comstock B, Bresnahan BW, Chen LE, Deyo RA, 2011;16(4):378–83. Halabi S, Turner JA, Avins AL, James K, et al. Systematic literature review of 38. Mhuiris AN, Volken T, Elliott JM, Hoggarth M, Samartzis D, Crawford RJ. imaging features of spinal degeneration in asymptomatic populations. Am J Reliability of quantifying the spatial distribution of fatty infiltration in lumbar Neuroradiol. 2015;36(4):811–6. paravertebral muscles using a new segmentation method for T1-weighted 17. Crawford RJ, Volken T, Valentin S, Melloh M, Elliott J. Rate of lumbar MRI. BMC Musculoskelet Disord. 2016;17(1):234. paravertebral muscle fat infiltration versus spinal degeneration in 39. Putzier M, Hartwig T, Hoff EK, Streitparth F, Strube P. Minimally invasive TLIF asymptomatic populations: an age- aggregated cross-sectional simulation leads to increased muscle sparing of the multifidus muscle but not the study. BMC Scoliosis Spinal Disord. 2016;11(1):21. longissimus muscle compared with conventional PLIF-a prospective 18. Fortin M, Yuan Y, Battie MC. Factors associated with paraspinal muscle randomized clinical trial. Spine Journal. 2016;16(7):811–9. asymmetry in size and composition in a general population sample of men. 40. Valenzuela W, Ferguson SJ, Ignasiak D, Diserens G, Vermathen P, Boesch C, Phys Ther. 2013;93(11):1540–50. Reyes M. Correction tool for active shape model based lumbar muscle 19. Valentin S, Licka T, Elliott J. Age and side-related morphometric MRI segmentation. PLoS One. 2015;2015:3033–6. evaluation of trunk muscles in people without back pain. Man Ther. 2015; 41. Elliott JM, Walton DM, Rademaker A, Parrish T. Quantification of cervical 20(1):90–5. spine muscle fat: a comparison between T1-weighted and multi-echo 20. Valentin S, Licka TF, Elliott J. MRI-determined lumbar muscle morphometry gradient echo imaging using a variable projection algorithm (VARPRO). BMC in man and sheep: potential biomechanical implications for ovine model to Med Imaging. 2013;11:13–30. https://doi.org/10.1186/1471-2342-13-30. human spine translation. J Anat. 2015;227(4):506–13. 42. Crawford RJ, Cornwall J, Abbott R, Elliott J. Manually defining regions of 21. Hancock M, Maher C, Macaskill P, Latimer J, Kos W, Pik J. MRI findings are interest when quantifying paravertebral muscles fatty infiltration from axial more common in selected patients with acute low back pain than controls? magnetic resonance imaging: a proposed method for the lumbar spine Eur Spine J. 2012;21(2):240–6. with anatomical cross-reference. BMC Musculoskelet Disord. 2017:18(25). Elliott et al. BMC Musculoskeletal Disorders (2018) 19:171 Page 14 of 14 43. Fleckenstein JL, Watamull D, Conner KE, Ezaki M, Greenlee RG Jr, Bryan WW, 66. Valentin S, Yeates TD, Licka T, Elliott J. Inter-rater reliability of trunk muscle Chason DP, Parkey RW, Peshock RM, Purdy PD. Denervated human skeletal morphometric analysis. J Back Musculoskeletal Rehabil. 2015;28(1):181–90. muscle: MR imaging evaluation. Radiology. 1993;187(1):213–8. 67. Abbott R, Peolsson A, West J, Elliott JM, Aslund U, Karlsson A, Dahlqvist 44. Fritz RC, Domroese ME, Carter GT. Physiological and anatomical basis of Leinhard O: The qualitative grading of muscle fat infiltration in whiplash muscle magnetic resonance imaging. Phys Med Rehabil Clin N Am. 2005; using fat/water magnetic resonance imaging. Spine J 2017, Sep 5. pii: 16(4):1033–51. x S1529–9430(17)30907–5. https://doi.org/10.1016/j.spinee.2017.08.233. [Epub ahead of print]. 45. Wokke BH, Bos C, Reijnierse M, van Rijswijk CS, Eggers H, Webb A, Verschuuren JJ, Kan HE. Comparison of Dixon and T1-weighted MR 68. Elliott JM, Pedler AR, Jull GA, Van Wyk L, Galloway GG, O'Leary SP. Differential changes in muscle composition exist in traumatic and methods to assess the degree of fat infiltration in duchenne muscular nontraumatic neck pain. Spine. 2014;39(1):39–47. dystrophy patients. J Magn Reson Imaging. 2013;38(3):619–24. 69. Boom HP, van Spronsen PH, van Ginkel FC, van Schijndel RA, Castelijns JA, 46. Elliott JM, Galloway GJ, Jull GA, Noteboom JT, Centeno CJ, Gibbon WW. Tuinzing DB. A comparison of human jaw muscle cross-sectional area and volume Magnetic resonance imaging analysis of the upper cervical spine extensor in long- and short-face subjects, using MRI. Arch Oral Biol. 2008;53(3):273–81. musculature in an asymptomatic cohort: an index of fat within muscle. Clin 70. Abbott R, Pedler A, Sterling M, Hides J, Murphey T, Hoggarth M, Radiol. 2005;60(3):355–63. Elliott J. The geography of fatty infiltrates within the cervical 47. Elliott JM, Jull GA, Noteboom JT, Durbridge GL, Gibbon WW. Magnetic multifidus and semispinalis cervicis in individuals with chronic resonance imaging study of cross-sectional area of the cervical extensor whiplash-associated disorders. J Orthop Sports Phys Ther. 2015;45(4):8. musculature in an asymptomatic cohort. Clin Anat. 2007;20(1):35–40. 71. Elliott J. Are there implications for morphological changes in neck muscles after 48. Sinclair CD, Morrow JM, Miranda MA, Davagnanam I, Cowley PC, Mehta H, whiplash injury? Spine (Phila Pa 1976). 2011;1(36(25 Suppl)):S205–10. Review Hanna MG, Koltzenburg M, Yousry TA, Reilly MM, et al. Skeletal muscle MRI 72. Bley TA, Wieben O, Francois CJ, Brittain JH, Reeder SB. Fat and water magnetisation transfer ratio reflects clinical severity in peripheral magnetic resonance imaging. J Magn Reson Imaging. 2010;31(1):4–18. neuropathies. J Neurol Neurosurg Psychiatry. 2012;83(1):29–32. 73. Reeder SB, McKenzie CA, Pineda AR, Yu H, Shimakawa A, Brau AC, 49. Sinclair CD, Morrow JM, Janiczek RL, Evans MR, Rawah E, Shah S, Hanna MG, Hargreaves BA, Gold GE, Brittain JH. Water-fat separation with IDEAL Reilly MM, Yousry TA, Thornton J. Stability and sensitivity of water T2 gradient-echo imaging. J Magn Reson Imaging. 2007;25(3):644–52. obtained with IDEAL-CPMG in healthy and fat-infiltrated skeletal muscle. 74. Costa DN, Pedrosa I, McKenzie C, Reeder SB, Rofsky NM. Body MRI using NMR Biomed. 2016;29(12):1800–12. IDEAL. Am J Roentgenol. 2008;190(4):1076–84. 50. Haase A, Frahm J, Hänicke W, Matthaei D. 1H NMR chemical shift selective 75. Gerdes CM, Kijowski R, Reeder SB. IDEAL imaging of the musculoskeletal (CHESS) imaging. Phys Med Biol. 1985;30:341–4. system: robust water fat separation for uniform fat suppression, marrow 51. Kaale BR, Krakenes J, Albrektsen G, Wester K. Whiplash-associated disorders evaluation, and cartilage imaging. AJR Am J Roentgenol. 2007;189(5):W284–91. impairment rating: neck disability index score according to severity of MRI 76. Romu T, Dahlqvist Leinhard O, Dahlström N, Borga M. Robust water fat findings of ligaments and membranes in the upper cervical spine. J separated dual-Echo MRI by phase-sensitive reconstruction. Magn Reson Neurotrauma. 2005;22(4):466–75. Med. 2017;78(3):1208–16. 52. Krakenes J, Kaale BR. Magnetic resonance imaging assessment of 77. Gerdle B, Forsgren MF, Bengtsson A, Leinhard OD, Soren B, Karlsson A, craniovertebral ligaments and membranes after whiplash trauma. Spine. Brandejsky V, Lund E, Lundberg P. Decreased muscle concentrations of ATP 2006;31(24):2820–6. and PCR in the quadriceps muscle of fibromyalgia patients - a (31) P-MRS 53. Krakenes J, Kaale BR, Moen G, Nordli H, Gilhus NE, Rorvik J. MRI assessment study. Eur J Pain. 2013;17(8):1205–15. of the alar ligaments in the late stage of whiplash injury–a study of 78. Elliott J, Sterling M, Noteboom JT, Treleaven J, Galloway G, Jull G. The structural abnormalities and observer agreement. Neuroradiology. 2002; clinical presentation of chronic whiplash and the relationship to findings of 44(7):617–24. MRI fatty infiltrates in the cervical extensor musculature: a preliminary 54. Myran R, Kvistad KA, Nygaard OP, Andresen H, Folvik M, Zwart JA. Magnetic investigation. Eur Spine J. 2009;18(9):1371–8. resonance imaging assessment of the alar ligaments in whiplash injuries: a 79. Fernandez-de-Las-Penas C, Bueno A, Ferrando J, Elliott JM, Cuadrado ML, Pareja case-control study. Spine. 2008;33(18):2012–6. JA. Magnetic resonance imaging study of the morphometry of cervical extensor 55. Ronnen HR, de Korte PJ, Brink PR, van der Bijl HJ, Tonino AJ, Franke C. Acute muscles in chronic tension-type headache. Cephalalgia. 2007;27(4):355–62. whiplash injury: is there a role for MR imaging?–a prospective study of 100 80. Okada E, Matsumoto M, Ichihara D, Chiba K, Toyama Y, Fujiwara H, Momoshima patients. Radiology. 1996;201(1):93–6. S, Nishiwaki Y, Takahata T. Cross-sectional area of posterior extensor muscles of 56. Cornwall J, Farrell SF, Sheard P. Fibre types of human suboccipital muscles. the cervical spine in asymptomatic subjects: a 10-year longitudinal magnetic Eur J Anat. 2016;20(1):31–6. resonance imaging study. Eur Spine J. 2011;20(9):1567–73. 57. Au J, Perriman DM, Pickering MR, Buirski G, Smith PN, Webb AL. Magnetic 81. Fortin M, Dobrescu O, Jarzem P, Ouellet J, Weber MH. Quantitative resonance imaging atlas of the cervical spine musculature. Clin Anat. 2016; magnetic resonance imaging analysis of the cervical spine extensor 29(5):643–59. muscles: Intrarater and interrater reliability of a novice and an experienced 58. Farrell SF, Osmotherly PG, Cornwall J, Sterling M, Rivett D. Cervical spine rater. Asian Spine J. 2018;12(1):94–102. meniscoids: an update on their morphological characteristics and potential 82. Inoue H, Montgomery S, Aghdasi B, Tan Y, Tian H, Jian X, Terrell R, Singh V, clinical significance. Eur Spine J. 2017;26(4):939–47. Wang J. Analysis of relationship between Paraspinal muscle fatty 59. Kennedy E, Albert M, Nicholson H. The fascicular anatomy and peak force degeneration and cervical spine motion using kinetic magnetic resonance capabilities of the sternocleidomastoid muscle. Surg Radiol Anat. 2017;39(6): imaging. Glob Spine J. 2012;2(1):33–8. 629–45. 83. Mitsutake T, Sakamoto M, Chyuda Y, Oka S, Hirata H, Matsuo T, Oishi T, 60. Cornwall J, Kennedy E. Fiber types of the anterior and lateral cervical Horikawa E. Greater cervical muscle fat infiltration evaluated by magnetic muscles in elderly males. Eur Spine J. 2015;24(9):1986–91. resonance imaging is associated with poor postural stability in patients with 61. Miller A, Woodley SJ, Cornwall J. Fibre type composition of female longus cervical Spondylotic radiculopathy. Spine. 2016;41:1. capitis and longus colli muscles. Anat Sci Int. 2016;91(2):163–8. 84. Choi MK, Kim SB, Park CK, Lee SH, Jo DJ. Relation of deep Paraspinal 62. Cornwall J, Stringer MD, Duxson M. Functional morphology of the thoracolumbar Muscles' cross-sectional area of the cervical spine and bone Union in transversospinal muscles. Spine (Phila Pa 1976). 2011;36(16):E1053–61. Anterior Cervical Decompression and Fusion: a retrospective study. World 63. Standring S, Anand N, Birch R, Collins P, Crossman AR, Gleeson M, et al: Neurosurg. 2016;96:91–100. Gray's anatomy : the anatomical basis of clinical practice., 41st edn. New 85. Cagnie B, Barbe T, Vandemaele P, Achten E, Cambier D, Danneels L. MRI York: Elsevier; 2016. analysis of muscle/fat index of the superficial and deep neck muscles in an 64. Smith AC, Parrish TB, Hoggarth MA, McPherson JG, Tysseling VM, asymptomatic cohort. Eur Spine J. 2009;18(5):704–9. Wasielewski M, Kim H, Hornby TG, Elliott J. Potential associations between 86. Uthaikhup S, Assapun J, Kothan S, Watcharasaksilp K, Elliott JM. Structural chronic whiplash and incomplete spinal cord injury. Spinal Cord Ser Cases. changes of the cervical muscles in elder women with cervicogenic 2015;2015:15024. https://doi.org/10.1038/scsandc.2015.24. headache. Musculoskelet Sci Pract. 2017;29:1–6. 65. Reeder SB, Hu HH, Sirlin CB. Proton density fat-fraction: a standardized mr-based biomarker of tissue fat concentration. J Magn Reson Imaging. 2012;36(5):1011–4.

Journal

BMC Musculoskeletal DisordersSpringer Journals

Published: May 28, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off