Evaluation of gold fiducial marker manual localisation for magnetic resonance-only prostate radiotherapy

Evaluation of gold fiducial marker manual localisation for magnetic resonance-only prostate... Background: The use of intraprostatic gold fiducial markers (FMs) ensures highly accurate and precise image-guided radiation therapy for patients diagnosed with prostate cancer thanks to the ease of localising FMs on photon-based imaging, like Computed Tomography (CT) images. Recently, Magnetic Resonance (MR)-only radiotherapy has been proposed to simplify the workflow and reduce possible systematic uncertainties. A critical, determining factor in the accuracy of such an MR-only simulation will be accurate FM localisation using solely MR images. Purpose: The aim of this study is to evaluate the performances of manual MR-based FM localisation within a clinical environment. Methods: We designed a study in which 5 clinically involved radiation therapy technicians (RTTs) independently localised the gold FMs implanted in 16 prostate cancer patients in two scenarios: employing a single MR sequence or a combination of sequences. Inter-observer precision and accuracy were assessed for the two scenarios for localisation in terms of 95% limit of agreement on single FMs (LoA)/ centre of mass (LoA ) and inter-marker CM distances (IDs), respectively. Results: The number of precisely located FMs (LoA<2 mm) increased from 38/48 to 45/48 FMs when localisation was performed using multiple sequences instead of single one. When performing localisation on multiple sequences, imprecise localisation of the FMs (3/48 FMs) occurred for 1/3 implanted FMs in three different patients. In terms of precision, we obtained LoA within 0.25 mm in all directions over the precisely located FMs. In terms of accuracy, IDs CM difference of manual MR-based localisation versus CT-based localisation was on average (±1STD)0.6±0.6 mm. Conclusions: For both the investigated scenarios, the results indicate that when FM classification was correct, the precision and accuracy are high and comparable to CT-based FM localisation. We found that use of multiple sequences led to better localisation performances compared with the use of single sequence. However, we observed that, due to the presence of calcification and motion, the risk of mislocated patient positioning is still too high to allow the sole use of manual FM localisation. Finally, strategies to possibly overcome the current challenges were proposed. Keywords: Magnetic resonance imaging, Radiotherapy treatment planning, MR-only treatment planning, Pre-treatment positioning, Fiducial marker localization, Manual detection, Accuracy, Precision *Correspondence: m.maspero@umcutrecht.nl; matteo.maspero.it@gmail.com Universitair Medisch Centrum Utrecht, Heidelberglaan 100, 3508 GA Utrecht, The Netherlands © 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. Maspero et al. Radiation Oncology (2018) 13:105 Page 2 of 12 Background use of MR images as reference for position verification The use of intraprostatic fiducial markers (FMs) ensures purposes [32]. highly accurate and precise image-guided radiation ther- Up to now, research on such an MR-only workflow has apy (IGRT) for patients diagnosed with prostate can- mainly focused upon generation of so-called synthetic- cer [1]. Specifically, it has been shown that markers CT images to allow dose calculation based on MR image enable for a safe reduction of PTV margin [2–5]. To information alone. Less attention has been paid to the accurately position patients based on the target loca- issue of MR-based FM localisation, which is a major deter- tion, a set of three non-co-linear markers centred in mining factor in obtaining accurate radiation treatment of the prostate is the minimum requirement allowing tri- prostate cancer. angulation and measurement of the prostate position in In MRI, FMs are depicted as signal voids in magnitude different planes [6]. Most markers are made of inert images since they do not produce nuclear magnetic res- metals (gold and titanium, for example). For prostate onance signal [33]. The appearance of FM voids varies cancer, gold FMs are generally employed and their use according to imaging parameters [34]and theFMorien- allows accurate prostate localisation [7]onphoton-based tation with respect to the magnetic field [35]. Up to now, images thanks to their high density, which increases radio to minimise manual interaction, automated MR-based FM opacity [8]. Markers have usually a cylindrical shape, a localisation methods have been proposed [36–39]. These diameter ranging between 0.5 and 1.5 mm and length methods are promising, resulting in acceptable accuracy between 2 and 10 mm; they remain in the patient per- and relatively high detection rates ranging from 84% to manently [7]. To ensure geometrically accurate IGRT 96%. However, we can not rely on the fact that all FMs will treatments, FMs are localised on computed tomography be automatically localised, leaving to manual observers (CT) during treatment simulation and they are localised the burden to correct for missed detections. These missed on kV/MV imaging before patient irradiation to verify detections can derive from misclassifying blood clots or and eventually match the pre-treatment patient position calcification as FMs since they all appear as signal voids in with the planned position [9, 10]. The localisation is MR images [7, 40]. Therefore, in addition to the require- performed based on a distinct image contrast the FMs ment of achieving a high localisation accuracy, the risk of induce in the CT as well as kV/MV images [11]. The misclassification of signal voids (i.e. false positives) should distinctness of the contrast ensures that false classifica- be very low as this could result in systematic errors in tion of FMs, e.g. due the presence of calcifications, is patient positioning [41]. unlikely. In current clinical practice, MR-based manual FM local- Recently, magnetic resonance imaging (MRI) and its isation is performed by radiation therapy technicians superior soft tissue contrast with respect to photon-based (RTTs) for registering MR to CT images [17, 41]. In such imaging [12, 13] enabled more accurate delineation of a setting, the presence of CT images greatly aids the RTT the prostate [14–16]. To exploit the advantages offered to discern whether signal voids in MR images can be by this imaging modality, the use of MRI in radiother- classified as FM, calcifications or blood clots. apy planning is rapidly expanding [6]. Nowadays, the While this MR-based FM localisation for CT/MR treatment simulation is generally based on CT and MRI simulation is clinically accepted, no previous study has images. Before treatment target delineation, CT and MR investigated the reliability of a solely MR-based FM images are rigidly registered based on the location of localisation by manual observers, which is the expected FMs on both image modalities [17, 18]. The accuracy of scenario within MR-only radiotherapy. the registration is generally considered as being within The aim of this study is to evaluate the performance of 1mm[19]. manual FM localisation during the planning of prostate More recently, MRI-based radiotherapy - also called cancer patients’ external beam radiotherapy treatments. “MR-only” radiotherapy - has been proposed [20–22] Furthermore, since in a clinical environment multiple to reduce systematic spatial uncertainties introduced sequences are usually available, we aim at investigat- when registering CT and MRI images [23]. More- ing whether the use of multiple sequences may impact over, an MR-only workflow would reduce costs of themanualFMlocalisation. We conductedastudyto the treatment and patient exposure to ionising radi- assess the inter-observer precision and accuracy of the ation [24]. Additionally, MR-only treatment planning MR-based FM localisation among RTTs. Furthermore, is particularly desirable in the context of MR-guided we evaluated in our patients’ group the occurrence of photon [25–27] and eventually proton [28–30] radio- misclassification of FMs. therapy. On the other hand, the introduction of an MR-only radiotherapy pathway raises a series of chal- Methods lenges as enabling MR-based dose calculations [22], This study is divided into three parts. First, we selected dealing with distortions in MR images [31], and the patients and acquired CT and MR images (Patient Maspero et al. Radiation Oncology (2018) 13:105 Page 3 of 12 preparation and selection section). Second, we per- MR-based fiducial marker localisation formed a multi-observer manual MR-based localisation Among the acquired MR images, we tested manual FM (MR-based fiducial marker localisation section). Finally, localisation on the three following sequences, for which we evaluated the precision and accuracy of the manual FM imaging parameters are reported in Table 1: localisation and investigated whether the observer agree- ment may lead to a precise and accurate patient alignment 1. a 3D Cartesian balanced steady-state free precession (Statistical analysis section). (bSSFP) sequence with spectral attenuated inversion recovery to obtain fat suppression and highlight Patient preparation and selection prostate boundaries. The images acquired with this The study was performed on patients with prostate carci- sequence were used by the physician to perform noma diagnosis who underwent radiotherapy planning at prostate delineations. The vendor’s name for this the University Medical Center Utrecht (The Netherlands) sequence was “3D balanced turbo field echo”. between September and October 2015. The study has 2. a 3D Cartesian T1-weighted dual radio frequency been conducted in accordance with regulations from the spoiled gradient-recalled echo (SPGR) sequence. The local ethical committee. SPGR sequence was acquired right after the bSSFP For position verification purposes, each patient received and used by the physician to distinguish bleedings three intraprostatic cylindrical gold FMs (HA2 Medizin- from the primary lesion. The vendor’s name for this technik GmbH, Germany) measuring 1 mm (diameter) by sequence was “3D T1 fast field echo”. 5 mm (length). The FMs were transperineally implanted 3. a 3D Cartesian dual gradient-recalled echo (GRE) under ultrasound guidance by a physician prior to the sequence. This sequence was acquired at the end of imaging session using two 18-gauge needles placed with the examination to have an independent sequence the aid of a template. for FM localisation. The field of view (FOV) was Patient positioning at CT scan (Brilliance CT Big Bore, reduced to the sole target reducing acquisition time Philips Medical Systems, Cleveland, Ohio, USA) was con- as well as making this sequence less prone to motion ducted simulating the treatment, i.e. using a flat table, artefacts. The vendor’s name for this sequence was knee wedges, positioning arms on the chest and tattooing “3D fast field echo”. the patient with the aid of laser alignment. Patient setup at 3T MR scan (Ingenia Omega HP, The sequences were selected as best candidates Philips Healthcare, Best, The Netherlands) was performed for FM identification after inspection of previously acquired patient images. The rationale underlying the using a knee wedge, but without a flat table top, with- out positioning arms on the chest and without laser- choice was the high contrast between FM location aided positioning. Patients were scanned using anterior and surrounding tissues and the spatial high resolution and posterior phased array coils (dS Torso and Posterior coils, 28 channels, Philips Healthcare, Best, The Nether- Table 1 Imaging parameters of the sequences used for gold FM lands). To avoid compression of patients’ anatomy, two manual localisation: the second column provides the details for in-house-built coil bridges supported the anterior coil. the balanced steady-state free precession (bSSFP) sequence, the The location of FMs from CT images was obtained as third column for the radio frequency spoiled gradient-recalled previously described in [39]. No rectum or bladder prepa- echo (SPGR) sequence acquired right after the bSSFP and the ration protocol was applied before imaging sessions, but fourth column for the gradient-recalled echo (GRE) sequence the patients were asked to empty their rectum in case rec- acquired at the end of the examination tal filling was noticed being larger than 5 cm during the Imaging parameters bSSFP SPGR GRE imaging session. TE /(TE )/TR [ms] 1.98/3.96 1.4/2.7/4.4 1.4/2.7/4.6 1 2 Criteria for selecting the subjects were: patients had Flip Angle [°] 40 10 10 gold FMs implanted prior to the imaging sessions, patients a 3 underwent CT and MRI on the same day acquiring FOV [mm ] 250x250x90 467x467x300 449x449x90 three specific MR sequences (see next section for fur- a Acquisition Matrix 252x234x90 312x314x200 376x376x75 ther details) and were without metallic implants. CT scans Reconstruction Matrix 512x512x90 320x320x200 400x400x75 were performed with the following imaging parameters: a 3 Reconstructed Voxel [mm ] 0.5x0.5x1.0 1.5x1.5x1.5 1.1x1.1x1.2 120 kV, exposure time = 923 ms, tube current between Bandwidth [Hz/voxel] 945 1078 1142 121 and 183 mA, in-plane matrix = 512x512 pixels, and Readout direction AP AP AP 3 mm slice thickness. The resolution was variable depend- ing on the field of view (FOV) used. The typical size of Acquisition time 4 min 29 s 4 min 1 s 2 min 10 s the FOV was 500x500x300 mm , which corresponds to an expressed in terms of anterior-posterior, right-left and superior-inferior directions in-plane resolution of 0.98x0.98 mm . The term FOV refers to field of view, AP to anterior-posterior Maspero et al. Radiation Oncology (2018) 13:105 Page 4 of 12 (voxel size <1.2x1.2x1.5 mm ). Furthermore, they were were identified, the FM located with the lowest reliabil- expected to preserve geometrical accuracy thanks to the ity was excluded from the statistical analysis. To generate 3D acquisition and large bandwidth (>900 Hz/pixel); sim- a consistent FM marker labelling among observers, FMs ilar sequences were used also by other institutions for FM were numbered (from 1 to 3) according to the position localisation [37, 42]. Five RTTs independently performed of their centre along the superior-inferior direction. In manual localisation of FMs by identifying the top and bot- case FMswerelocated in thesametransverseplane,the tom of the markers on magnitude images of the bSSFP left-right direction was used for labelling. To keep the con- sequence. The RTTs were requested to follow standard- sistency of the labelling among the observers, the labelling ised instructions regarding zoom and window/level of the was manually checked and, when necessary, corrected. images. The provided instructions are available as part The analysis was performed in Matlab (R2015a, the Math- of the Additional file 1. The RTTs involved in the study Works Inc., Natick, Massachusetts, United States). had varying experience as technicians: 11, 6.5, 5, 5 and 10 Inter-observer agreement or spatial precision years. They all had experience with position verification. Single FM locations For each observer, the number of Four RTTs (all the observers except the first observer), detected FMs was reported. An agreement position was had experience with image registration between MR and defined as the mean position among all the five observers. CT images of: 0, 2, 2, 2.5 and 2.5 years, respectively. They To assess the precision among the five observers, FM loca- were new to MR-only FM localisation since this procedure tions were compared calculating a 95% limit of agreement was initiated with this study. The observers were asked (= 1.96 times the standard deviation (STD) [43]) of the dis- to report for which patients the FM localisation was per- tance to the mean position (LoA) in the three directions ceived as being problematic and they got the freedom to (X = anterior-posterior, Y = right-left and Z = superior- provide three or four candidates. Also, the RTTs indicated inferior). For comparison with [44], a threshold for clinical which FM was the most difficult to distinguish. The FM acceptability was set to 95% LoA ≤ 2 mm. Bar plots centre was calculated as the mean position between the providing a visual assessment of the inter-observer vari- manually identified top and bottom positions. Note that ability were created reporting also a more stringent LoA the observed length of the FM may not correspond to threshold of 1 mm. An investigation of the CT and MR the nominal length of the FM (5mm). The term “appar- images was performed on an individual patient basis to ent” location was used to refer to the location of top and investigate the causes underlying imprecise localisation bottom as identified by the observers. Note that apparent of FMs; schematic representations of the inter-observer location of the top and bottom of each FM coincided with localisation were also examined. the centre of a voxel, thus, the calculated centre of the FMs may be located with resolution higher than a single voxel. Centre of mass locations To verify the impact on patient To investigate whether the use of multiple sequences alignment, the location of the centre of mass (CM) among impacts the FM localisation, the RTTs performed the all the FMs was calculated for each patient and observer. FM localisation again using multiple MR images after For each patient, the 95% limit of agreement (= 1.96 localisation using images of the sole bSSFP sequence. In times the standard deviation (STD) [43]) with the aver- particular, in this second localisation the RTTs employed age position of the CMs (LoA ) was calculated among CM images of the bSSFP, the second echo of the SPGR and all the observers in the three directions [44]. To assess the GRE sequences. Note that the location of the mark- agreement of the CM position among the observers a ers on the bSSFP was not made available when repeating threshold of LoA < 2mmwas used forcomparison CM the localisation. In the case of inter-scan FM motion, the with [44]. To verify clinical CM agreement, the threshold RTTs were instructed to consider the position of the FMs LoA < 1 mm was employed. Bar plots providing a CM on the bSSFP as a reference. visual assessment of the inter-observer agreement vari- For completeness, the following metrics were also ability were created. Note that patient alignment is gen- recorded: the apparent length as characterised in each erally performed on the centre of mass location [41]; sequenceand by thedifferent observers, thetimerequired therefore, this is considered as the final metric to assess by each observer to complete the FM localisation and the inter-observer precision. number of FMs for which the localisation was perceived as being problematic. These results are reported as part of Intra-observer agreement the Additional file 2. To evaluate whether, over all the observers, a statistically significant variation of FM location occurred between Statistical analysis localisation using only the bSSFP sequence and the combi- The analysis was performed on the FM centres as located nation of bSSFP, SPGR and GRE sequences, we performed using both a “single” sequence (bSSFP) and “multiple” Wilcoxon rank-sum test at the confidence level of 95% in MRI sequences (bSSFP, SPGR and GRE). In case four FMs the three directions on LoA and LoA . CM Maspero et al. Radiation Oncology (2018) 13:105 Page 5 of 12 Spatial accuracy x 5 observers x 2 sequence scenarios) performed over all The difference of inter-marker distances (IDs) between the observers. The FMs were implanted at least one week the FMs located in CT and MRI were calculated for the prior to imaging. precisely located (LoA < 2 mm) markers using single During the pre-planning imaging session, MRI scans and multiple sequences. For each observer and over all were performed within maximum 70 min (mean time the observers, the absolute difference between the ID of = 45 min, minimum time of 20 min and IQR = 34- MRI and CT were calculated as in [38] and characterised 50 min) after the CT scans. For all the patients, the in terms of mean, median, standard deviation (STD) and bSSFP and SPGR sequences were acquired one after each range ([minimum, maximum]). other with a maximum time difference of 5 min, while the GRE sequence was acquired at least 15 min after Results the SPGR sequence. Figure 1 shows a zoomed axial slice Seventeen consecutive prostate patients (61.4-81.9 years, of CT, bSSFP, SPGR and GRE images for a patient in mean age = 68.7years,medianage = 68.3 years, inter which all observers agreed on the locations of the FMs quartile range (IQR) = 66.1-70.8 years) were considered (P1, top) and a patient in which one FM was challenging forinclusion in thestudy.All thepatientswerestagedas (P9, bottom) T1c-3b, Gleason score ≥ 6 and one of the patients (P14) had a hip implant and was excluded from the analysis (the Single FM locations localisation and images for this patient are presented in All the observers detected three FMs for all the patients, the Additional file 3 in Fig. 3)). Within the patient popula- except one observer (Obs1) who detected four FMs for tion, the average prostate volume during imaging sessions two patients (P4 and P17) when using the bSSFP sequence. was 56.8 ml (range = 32.1-117.3 ml, median volume = When also the SPGR and GRE sequences were employed, 54.8 ml, IQR = 42.9-70.9 ml) and body mass index was all the RTTs localised three FMs. Figure 2 provides a 2 2 on average 26.4 kg/m (range = 19.9-30.7 kg/m ,IQR = schematic representation of the centres of the FMs as 24.9-28.7 kg/m ). No patient received adjuvant hormonal localised by all the observers for patients P1 and P9 using therapy. multiple sequences. The agreement position is as well Patients underwent intensity-modulated radiotherapy, shown. As example, taking into consideration the FM with using5beamsof10MV, with aprescribeddoseof77Gy the largest spread for patient P1 (FM 1), the LoA were 0.69, to theentireprostatein35fractions (2.2 Gy perfraction). 0.57, 0.84 mm in X (anterior-posterior), Y (right-left) and Other clinical prescriptions are specified in [45]. Each of Z (superior-inferior) directions, respectively. the patients had three FMs implanted leading to a total The bar plots of LoA for all the patients over the of 48 FMs and 580 FM localisations (16 patients x 3 FMs five RTTs is shown in Fig. 3.The LoAwas foundto Fig. 1 Zoom of an axial slice of CT (left), bSSFP (centre-left), SPGR (centre-right) and GRE (right) images for patients P1 (top) and P9 (bottom) before image registration. The axes X and Y indicate the anterior-posterior and right-left directions. The intensity of CT image is in Hounsfield Units (HU), while of MR images is normalised to the maximum over the whole dataset. Note the presence of calcifications for the patient P9; they are visible as high intensity on CT and signal void on MR images Maspero et al. Radiation Oncology (2018) 13:105 Page 6 of 12 Fig. 2 Schematic representation of the centres of the FMs as localised by all the observers for patient P1 (top) and P9 (bottom) using multiple sequences (bSSFP, SPGR and GRE). The labelling of the FM is indicated by the marker: ◦,x,+and • for FM having number 1, 2, 3 and for agreement position, respectively. For patient P1 (top), the LoA of FM1 over the five RTTs was 0.27, 0.31 and 0.38 mm in X, Y and Z, respectively, which was CM considered clinically acceptable; for patient P9 (bottom), the LoA for FM2 over the five observers was 11.17, 0.99 and 13.70 mm in X, Y and Z, CM respectively, which was considered clinically unacceptable be higher than 2 mm in one of the three directions was found to be > 2 mm for more than one FM per for 10/48 and 3/48 FMs when the observers located on patient, while when localisation was performed on mul- a single (bSSFP) and multiple (bSSFP, SPGR and GRE) tiple sequences LoA was > 2 mm only for one FM per sequences, respectively. This resulted in an increased patient. Focusing on the scenario with the largest agree- agreement (45/48) when the observers located on multi- ment (localisation performed using multiple sequences), ple sequences with respect to a single sequence (38/48). localisation of maximum one FM was found to be impre- Over all the three directions, the Wilcoxon rank-sum test cise for 3/16 patients: P4, P6 and P9. Excluding these at 95% confidence interval resulted in significantly differ- 3 FMs (considering, therefore, 45/48 FMs), the average ent LoA when comparing FM location obtained with one (±1STD)LoA was0.19 ±0.15, 0.18 ±0.12 and 0.30 or multiple sequences. In particular, as shown in Fig. 3 ±0.31 mm in anterior-posterior, right-left and superior- with the use of multiple sequences for two patients (P7 inferior directions, respectively. After the investigation of and P17) the LoA decreased below 2 mm. When local- the images acquired for the patients resulting in an impre- isingusing asinglesequenceoverall thepatients, LoA cise FM localisation, we observed that patients P4 (Fig. 1 Maspero et al. Radiation Oncology (2018) 13:105 Page 7 of 12 Fig. 3 The 95% limit of agreement (LoA) calculated, for each patient, over the five observers for the single fiducial marker (FM) in the three directions, where X = anterior-posterior (top), Y = right-left (center) and Z = superior-inferior (bottom). On the left is shown the FM localisation as performed using the bSSFP sequence only, while on the right the FM localisation as performed using multiple sequences. The dotted and dashed lines represent the LoA of 2 mm, while the dotted lines represent the LoA of 1 mm. Note that patient P14 had hip implant and the results are here presented but were excluded in the statistical analysis in Additional file 3) and P9 (Figs. 1 and 2 bottom) were bSSFP impacting reliability of the localisation for this FM. characterised by the presence of large (>2mmindiame- Considering the results from a different perspective, for ter) intra-prostatic calcifications. In both cases, 1/5 RTTs the total 240 (16 patients x 5 observers x 3 FMs) single (Obs1 for P4 and Obs3 for P9) localised one of the FMs observer localisations using multiple sequences, 2 times far away from the other four observers. Figure 2 shows calcifications were marked as FMs by one of the RTTs that a misclassification occurred for patient P9 when con- and no agreement could be found for one FM among all sidering the FM2 and Obs3. The same occurred for FM2 the five RTTs. This would result in misclassification for and Obs1 for patient P4 (Fig. 1 in Additional file 3 bot- 7 out of the 240 single observer localisations, or 7/80 (16 tom). After observing the location of the misclassified patients x 5 observers) single observer localisations of the FMsasreportedbythe twoobservers in theMRand CT CM in the case the outliers cannot be eliminated. images, we found that the FMs were located in corre- spondence of calcifications. For one patient (P6, as shown Center of mass locations in Additional file 3 in Fig. 2), one of the FMs was not visible Figure 4 presents the bar plot of the 95% LoA for CM on bSFFP but appeared on SPGR and GRE; we hypothe- all the patients over the five RTTs. The LoA was CM sised that motion reduced the visibility of the FM on the found to be >1 mm in one of the three directions (X, Fig. 4 The 95% limit of agreement of the centre of mass (LoA ) calculated, for each patient, over the five observers for a single fiducial marker (FM) CM in the three directions, where X = anterior-posterior (blue), Y = right-left (red) and Z = superior-inferior (green). On the left is shown the localisation of the CM as performed using the bSSFP sequence only, while on the right the localisation of the CM as performed using multiple sequences. The dotted and dashed lines represent the LoA of 2 mm, while the dotted lines represent the LoA of 1 mm. Note that patient P14 had hip implant and the results are here presented but were excluded in the statistical analysis Maspero et al. Radiation Oncology (2018) 13:105 Page 8 of 12 Y or Z) for 5/16 and 3/16 patients when the observers Note that this fundamentally differs from localisation of located on a single (bSSFP) and multiple (bSSFP, SPGR FMsonMRimagesinthe currentCT-MR simulation and GRE) sequences, respectively. Over all the three workflow for registration purposes as the CT images directions, the Wilcoxon rank-sum test at 95% confi- can be used to minimise misclassification on the MR dence interval resulted in significantly different LoA when images. In this sense, this study was conducted to verify comparing FM locations obtained with one or multiple whether an MR-only simulation could facilitate a robust sequences. Excluding the imprecisely located CMs, the positioning workflow comparable to current CT-based average (±1STD)LoA when localisation was per- positioning in all the cases. CM formed with single sequence was 0.10 ±0.05, 0.10 ±0.06 In this study, the use of multiple sequences led to pre- and 0.19 ±0.13 mm in anterior-posterior, right-left and cise localisation (LoA< 2 mm) in more patients and of superior-inferior directions, respectively; the average (±1 more FMs (13/16 patients and 45/48 FMs) than locali- STD) LoA when localisation was performed with mul- sation with a single sequence only (11/16 patients and CM tiple sequences was 0.11 ±0.06, 0.13 ±0.09 and 0.23 ±0.18 38/48 FMs). For both scenarios, the precision calculated mm in anterior-posterior, right-left and superior-inferior as the average of LoA in all the directions on the pre- CM directions, respectively. In all the directions, the average cise localised FMs was within 0.25 mm. The results are LoA is <0.25 mm. in good agreement with others [18, 44, 46, 47]. Huisman CM et al. [18] obtained a precision of 0.5 mm in the centroid Spatial accuracy of the prostate on a cohort of 21 patients when assess- Table 2 shows the mean, median, STD and range of the ing registration of CT and MR images. Ullman et al. [46] absolute difference in the ID of the precisely located reported a mean inter-observer variability of 0.9±0.6 mm FMs (LoA <2 mm) using single and multiple sequences. when performing registration on photon-based portal Among all the observers, the average ID difference is images. Deegan et al. [47] reported that inter-observer slightly lower (0.5±0.6 mm) when locating with mul- LoA on the applied registration, which is comparable to tiple sequences with respect to with a single sequence the LoA , was in the range of about ±2 mm. Litera- CM (0.7±0.6 mm). ture reporting single FM localisation precision has not been found. Discussion In general, in our study, when FMs were precisely The precision and accuracy of manual localisation of localised, they were also accurately localised. In particu- intraprostatic gold FMs using solely MR images was eval- lar, we found an inter-observer accuracy of 0.7 mm with the single sequence and of 0.6 mm with the multiple uated in the context of an MRI-only simulation workflow. sequences. These results are slightly more accurate than what presented when comparing a human observatory to Table 2 The mean, median, standard deviation (STD) and range automatic FM localisation by Gustafsson et al. [38]and ([min, max]) of the absolute difference in the inter-marker in line with the accuracy previously considered acceptable distances (IDs) of the precisely located FMs between CT and MRI for the CM localisation performed with photon-based for all the single observers and for all the five observers imaging (0.6 mm) [48]. Sequence Observer Mean Median STD Range However, for a single FM in 3/16 cases, precise local- isation was not achieved. That implies that a correct Single 1 0.8 0.6 0.7 [0.1, 3.1] positioning in these patients can not be guaranteed. Based 2 0.6 0.5 0.5 [0.0, 2.5] on the thorough investigation of the images of these spe- 3 0.7 0.6 0.5 [0.0, 2.1] cific patients, we concluded that the following two causes 4 0.7 0.6 0.6 [0.1, 2.9] may have led to imprecise FM localisation: (1) presence of 5 0.7 0.5 0.6 [0.1, 2.5] calcifications miscalssified as FM and (2) motion during All 0.7 0.6 0.6 [0.0, 3.1] the bSSFP sequence. (1) Previous studies reported the presence of calcifications in 40 to 88% of prostate cancer Multiple 1 0.7 0.4 0.6 [0.0, 2.7] patients [7, 38, 40, 49]. In our study, for 2/16 patients the 2 0.6 0.4 0.6 [0.0, 3.0] presence of calcifications led to misclassified FM localisa- 3 0.7 0.5 0.7 [0.0, 2.5] tion for 1/5 RTT. Interestingly, the observers seemed to be aware of the difficulties and they reported that the local- 4 0.7 0.5 0.6 [0.0, 2.8] isation procedure for such patients was problematic (see 5 0.7 0.6 0.5 [0.0, 2.5] Fig. 1 in the Additional file 2). (2) Motion as a possible All 0.6 0.5 0.6 [0.0, 3.0] cause of hampered accuracy of FM localisation has already The results were calculated excluding 10/48 and 3/48 FMs for the localisation been reported in the literature for the bSSFP sequence performed on a single and multiple sequences, respectively. All the values are [42]. The readout of this sequence was 3D leading to expressed in mm Maspero et al. Radiation Oncology (2018) 13:105 Page 9 of 12 typical acquisition times of 2–3 min, and thus motion visualisation and also manual localisation performance. 2) blurring is likely to occur. Among the available MRI sequences, only the images of To obtain accurate localisation for all the patient cases, one echo of the gradient-echo sequences have been taken we believe that redundancy should be added in the locali- into consideration in this study. It may occur that acquir- sation procedure to lower the risk of FM misclassification. ing with different image parameters or MR sequences In this sense, we foresee the following as possible ways to may result in more favourable manual FM localisation increase the redundancy: performance. For example, recently, the use of multi-echo images showed promising results, thanks to the increasing multiple observers localisation. Whenever an RTT size of a signal void when increasing the echo time [38]; would have low confidence in the FM localisation, an Future studies could investigate whether MR sequence independent observer could perform localisation and optimisation or the use of other sequences may be more assess a posteriori the initially found position. In this suitable for FM localisation, verifying accuracy and preci- scenario, the experience of the RTT may influence sion performance. the outcome. Further investigations are necessary to From a general perspective, in our study, five RTTs evaluate whether such scenario will lead to accurate were involved, making the findings representative of localisation in all the cases. a realistic situation. As the observers were not famil- implantation of a fourth marker. The use of a fourth iar with MR-only FM localisation, it may be expected FM could be easily performed without increasing the that better results may be obtained by training the patient discomfort: the fourth FM could be observers for this specific context. In this sense, it collinearly placed with the third FM avoiding a new may be interesting to verify, in a future study, the needle insertion. In case of FM misclasssification, the influence of clinical experience on manual localisation RTTs may explicitly exclude one of the FM when performance. correcting patient set-up, remaining with a sufficient Comparing our study to previous research, a limita- number of FMs to enable the procedure. On the tion of the presented cohort is its size, although, no other other hand, with four FMs several permutations of 3 research has been presented to assess manual FM locali- FMs could be considered and the RTTs would need sation with such details and reporting localisation perfor- consistently choose the FMs between imaging mances within a realistic clinical environment. Recently, modalities to obtain identical set-up corrections. Gustafsson et al. [38] presented results of the accu- resorting to automatic localisation. Given the racy of manual FM localisation and a larger cohort promising result obtained with automatic gold FM (44 patients). Unfortunately, the precision has not been localisation methods [36–39], resorting to a reported. combination of manual and automated MR-based Recently, the use of MR-visible fiducial markers have FM localisation methods may ensure safe MR-based been proposed offering new possibilities for MR-based simulation of patient position. marker localisation [34, 50]. In addition, FM localisation In our institution further investigation is ongoing to may also be based on mechanisms other than imaging. verify that using automatic localisation [39]isaviable For example, it has been shown that transponders can approach including also the insertion of a fourth FM. be safely implanted ensuring real-time prostate localisa- Alternatively to the redundancy options above proposed, tion [51]. Both these approaches may be adopted in an another centre [42] reported that using kV radiogra- MR-only workflow offering an alternative to gold FM phy after FM implantation provided independent images localisation. that facilitated MR-based FM localisation. Similarly to In the perspective of MR-only Radiotherapy, and con- this approach, we could also speculate about designing sidering the case of gold FM, the use of multiple sequences a workflow that foresees referring the patients to CT would enable manual marker localisation for precise and in case of dubious manual FM localisation at the MR accurate simulation of prostate cancer patients’ position scan. Performing a low dose CT for all the patients prior to irradiation in almost all the cases. Nevertheless, for the sole purpose of FM localisation could be believing that an MR-only simulation should facilitate a another possibility. robust positioning workflow, we think the risk of mislo- Strategies to possibly solve FM misclassification other cated patient positioning is still too high and that addi- than adding redundancy may involve 1) further MR tional redundancy is essential to enable a safe clinical sequence optimisation and 2) employing different MR practice. sequences. 1) Further MR sequence optimisation could, for example, be employed to diminish the susceptibility to Conclusion motion by reducing the acquisition time of the employed We studied inter-observer precision and accuracy of man- sequences. In addition, sequence optimisation may impact ual gold FM localisation for MR-only prostate cancer Maspero et al. Radiation Oncology (2018) 13:105 Page 10 of 12 external beam therapy simulation over five RTTs for two observers (bottom) are shown in Figs. 1 and 2, respectively. For completeness, we report also the CT ad MRI images along with the scenarios: employing a single MRI sequence (bSSFP) or schematic representation of the centres of the FM for patient P14 in Fig. 3. a combination of multiple sequences (bSSFP, SPGR and Note that this patient was not considered during the analysis since had a GRE). The use of multiple sequences (bSSFP, SPGR and hip implant. (PDF 1823 kb) GRE) led to better localisation performances compared with the use of a single sequence (bSSFP). For both the Abbreviations scenarios, the results indicate that when FM classification AP: Anterior-posterior; bSSFP: Balanced steady-state free precession; CT: Computed tomography; CM: Centre of mass; FM: Fiducial marker; FOV: Field of was correct, the precision and accuracy are high and com- view; GRE: Gradient-recalled echo; HU: Hounsfield units; IGRT: Image-guided parable to CT-based FM localisation. However, the risk radiotherapy; IQR: Inter quartile range; kV: Kilo Volt; LoA: Limit of agreement; of mislocated patient positioning due to FM misclassifi- LoA : Limit of agreement of the centre of mass; MV: Mega volt; MR(I): CM Magnetic resonance (imaging); Obs: Observer; P: Patient; PTV: Planning target cation is still too high to allow the sole use of manual volume; RTT: Radiation therapy technician; SPGR: Spoiled gradient-recalled FM localisation. For future work, we hypothesise that fur- echo; STD: Standard deviation; TE: Echo time; TR: Repetition time ther increasing redundancy by increasing the number of Acknowledgements FM per patient and by setting up a system to rely on We are grateful to Nicole Vissers, Joske Boudewijn and Tiny Vlig (UMC Utrecht, multiple observations or automatic localisation is nec- The Netherlands) for performing the manual localisation and their kind essary to increase the detection rate and enable clinical collaboration. We would like to thank Gert J Meijer (UMC Utrecht, The Netherlands) for discussion about the design of the study and Max A Viergever introduction. (UMC Utrecht, The Netherlands) and Jan J W Lagendijk (UMC Utrecht, The Netherlands) for providing general support to the research. Additional files Funding The research is funded by ZonMw IMDI Programme, project number: 1040030. The project is co-funded by Philips Healthcare. Additional file 1: Instructions provided to the clinical observers. As part of the supplementary material is possible to download a repository Availability of data and materials (InstructionPackage.zip) containing the instructions provided to the RTTs The datasets analysed during the current study are not publicly available due before performing the manual FM localisation. In particular, the repository to internal policy of the Medical Ethical Commission pertaining to data sharing contains the following files: but are available from the corresponding author upon Medical Ethical [1.] GeneralGuidelineFMloc.pdf which presents a short description of the Commission’s approval. The datasets generated during the inter-observer procedure; localisation, e.g. the FM location and the correspondent analysis are available [2.] PracticalInstructionFMloc.pdf which describes step-by-step the at https://matteomaspero.github.io/Manual-gold-FM-localisation/. procedure; [3.] Checklist_Obs.pdf which is aimed at supporting the RTTs during the Authors’ contributions procedure in keeping track and annotate for which patient the localisation MM designed and managed the study, collected and analysed the data and was found problematic. (ZIP 194 kb) drafted/revised the manuscript. PRS participated to the study design and the revision of the manuscript. NJW, GGS and GJK participated to design the study, Additional file 2: Annotations on the FM localisation. As part of the to collect inter-observer FM localisations and revise the manuscript. HCJdB supplementary material, we report the apparent length of the FMs for each and JRNvdVvZ contributed to design the study and revise the manuscript. observer and the time spent by each observer performing the FM CATvdB participated to design the study and to revise the manuscript. All localisation over all the patients. In particular, Table 1 shows the mean, authors read and approved the final manuscript. standard deviation (STD), range [min, max] of the apparent length, expressed in mm. The weighted mean over all the observer is 7.5 ± 0.6 mm Ethics approval and consent to participate and 7.7 ± 0.7 mm for localisation using a single and multiple sequences, The study received approval of the medical ethical commission (Medisch respectively. Note that the apparent length was longer than the nominal Ethische Toetsingscommissie) and was classified under the protocol number length of the FM (5 mm). Table 2 reports the mean, STD and range of the 15-444/C approved on 29th July 2015. time needed by each observer to perform the FM localisation using single and multiple sequences. The weighted mean over all the observer is Competing interests 5.8 ± 1.4 min. Note that all the RTTs localised the FMs first using a single Peter R Seevinck declares to be a majority shareholder of MRIGuidance B.V. and then multiple sequences for all the patients. The RTTs were free to Cornelis A T van den Berg declares to be a minority shareholder of MRCode B.V. chose the order of patients and whether concluding the procedure first for each the patients using both single and multiple sequences or first for all Publisher’s Note the patients using single sequence and then repeat for all the patients Springer Nature remains neutral with regard to jurisdictional claims in using multiple modalities. Possible differences in the way the RTTs published maps and institutional affiliations. performed the procedure does not permit to understand whether the FM localisation is faster using single or multiple sequences. In addition, a Received: 27 October 2017 Accepted: 13 April 2018 histogram reporting the frequency of unreliable FM localisation, as perceived by the RTTs is shown in Fig. 1 for four out of five observers; one of the observers did not report the reliability of the localisation. The observers References reported the perceived reliability without distinction between localisation 1. Zaorsky NG, Showalter TN, Ezzell GA, Nguyen PL, Assimos DG, performed employing a single and multiple sequences. (PDF 108 kb) D’Amico AV, Gottschalk AR, Gustafson GS, Keole SR, Liauw SL, Lloyd S, Additional file 3: Single patient investigation. As a supplementary McLaughlin PW, Movsas B, Prestidge BR, Taira AV, Vapiwala N, Davis BJ. material, we report CT and MRI images for the patients P4 and P6, which ACR appropriateness criteria® external beam radiation therapy treatment were found having LoA > 2 mm in maximum one of the three FMs for planning for clinically localized prostate cancer, Part II of II. Advances localisation performed with multiple sequences. Zoom of an axial slice of Radiat Oncol. 2017. https://doi.org/10.1016/j.adro.2017.03.003. CT (top left), bSSFP (top centre-left), SPGR (top centre-right) and GRE (top 2. Wu J, Haycocks T, Alasti H, Ottewell G, Middlemiss N, Abdolell M, right) images for the patients P4, P6 before image registration as well as Warde P, Toi A, Catton C. Positioning errors and prostate motion during schematic representations of the centres of the FMs as localised by all the conformal prostate radiotherapy using on-line isocentre set-up Maspero et al. Radiation Oncology (2018) 13:105 Page 11 of 12 verification and implanted prostate markers. Radiother Oncol. 2001;61(2): 20. Fraass BA, McShan DL, Diaz RF, Ten Haken RK, Aisen A, Gebarski S, 127–33. https://doi.org/10.1016/S0167-8140(01)00452-2. Glazer G, Lichter AS. Integration of magnetic resonance imaging into 3. Schallenkamp JM, Herman MG, Kruse JJ, Pisansky TM. Prostate position radiation therapy treatment planning: i. technical considerations. Int J relative to pelvic bony anatomy based on intraprostatic gold markers and Radiat Oncol Biol Phys. 1987;13(12):1897–908. https://doi.org/10.1016/ electronic portal imaging. Int J Radiat Oncol Biol Phys. 2005;63(3):800–11. 0360-3016(87)90358-0. https://doi.org/10.1016/j.ijrobp.2005.02.022. 21. Lee YK, Bollet M, Charles-Edwards G, Flower MA, Leach MO, McNair H, 4. Beltran C, Herman MG, Davis BJ. Planning Target Margin Calculations for Moore E, Rowbottom C, Webb S. Radiotherapy treatment planning of Prostate Radiotherapy Based on Intrafraction and Interfraction Motion prostate cancer using magnetic resonance imaging alone. Radiother Oncol. Using Four Localization Methods. Int J Radiat Oncol Biol Phys. 2008;70(1): 2003;66(2):203–16. https://doi.org/10.1016/S0167-8140(02)00440-1. 289–95. https://doi.org/10.1016/j.ijrobp.2007.08.040. 22. Edmund JM, Nyholm T. A review of substitute CT generation for MRI-only 5. Greer P, Dahl K, Ebert M, Wratten C, White M, Denham J. Comparison of radiation therapy. Radiat Oncol. 2017;12(1):28. https://doi.org/10.1186/ prostate set-up accuracy and margins with off-line bony anatomy s13014-016-0747-y. corrections and online implanted fiducial-based corrections. J Med 23. Nyholm T, Nyberg M, Karlsson MG. Systematisation of spatial Imaging Radiat Oncol. 2008;52(5):511–6. https://doi.org/10.1111/j.1440- uncertainties for comparison between a MR and a CT-based radiotherapy 1673.2008.02005.x. workflow for prostate treatments. Radiat Oncol. 2009;4(1):54. https://doi. 6. Schmidt MA, Payne GS. Radiotherapy planning using MRI. Phys Md Biol. org/10.1186/1748-717X-4-54. 2015;60(22):323–61. https://doi.org/10.1088/0031-9155/60/22/R323. 24. Karlsson M, Karlsson M. G, Nyholm T, Amies C, Zackrisson B. Dedicated 7. Ng M, Brown E, Williams A, Chao M, Lawrentschuk N, Chee R. Fiducial Magnetic Resonance Imaging in the Radiotherapy Clinic. Int J Radiat markers and spacers in prostate radiotherapy: current applications. BJU Oncol Biol Phys. 2009;74(2):644–51. https://doi.org/10.1016/j.ijrobp.2009. Int. 2014;113(S2):13–20. https://doi.org/10.1111/bju.12624. 01.065. 8. Gall K. P, Verhey L. J, Wagner M. Computer-assisted positioning of 25. Raaymakers BW, Raaijmakers AJE, Kotte ANTJ, Jette D, Lagendijk JJW. radiotherapy patients using implanted radiopaque fiducials. Med. Phys. Integrating a MRI scanner with a 6 MV radiotherapy accelerator: dose 1993;20(4):1153–9. https://doi.org/10.1118/1.596969. deposition in a transverse magnetic field. Phys Med Biol. 2004;49(17): 9. Balter JM, Lam KL, Sandler HM, Littles JF, Bree RL, Ten Haken RK. 4109–18. https://doi.org/10.1088/0031-9155/49/17/019. Automated localization of the prostate at the time of treatment using 26. Dempsey J, Benoit D, Fitzsimmons J, Haghighat A, Li J, Low D, Mutic S, implanted radiopaque markers: Technical feasibility. Int J Radiat Oncol Biol Palta J, Romeijn H, Sjoden G. A device for realtime 3D image-guided Phys. 1995;33(5):1281–6. https://doi.org/10.1016/0360-3016(95)02083-7. IMRT. Int J Radiat Oncol Biol Phys. 2005;63:202. 10. Vigneault E, Pouliot J, Laverdière J, Roy J, Dorion M. Electronic portal 27. Fallone BG, Murray B, Rathee S, Stanescu T, Steciw S, Vidakovic S, imaging device detection of radioopaque markers for the evaluation of Blosser E, Tymofichuk D. First MR images obtained during megavoltage prostate position during megavoltage irradiation: A clinical study. Int J photon irradiation from a prototype integrated linac-MR system. Med Radiat Oncol Biol Phys. 1997;37(1):205–12. https://doi.org/10.1016/S0360- Phys. 2009;36(6):2084–8. https://doi.org/10.1118/1.3125662. 3016(96)00341-0. 28. Raaymakers BW, Raaijmakers AJE, Lagendijk JJW. Feasibility of MRI 11. Habermehl D, Henkner K, Ecker S, Jäkel O, Debus J, Combs SE. guided proton therapy: magnetic field dose effects. Phys Med Biol. Evaluation of different fiducial markers for image-guided radiotherapy 2008;53(20):5615–22. https://doi.org/10.1088/0031-9155/53/20/003. and particle therapy. J. Radiat. Res. 2013;54(Suppl 1):61–8. https://doi.org/ 29. Moteabbed M, Schuemann J, Paganetti H. Dosimetric feasibility of 10.1093/jrr/rrt071. real-time MRI-guided proton therapy. Med Phys. 2014;41(11):111713. 12. Debois M, Oyen R, Maes F, Verswijvel G, Gatti G, Bosmans H, Feron M, https://doi.org/10.1118/1.4897570. Bellon E, Kutcher G, Van Poppel H, Vanuytsel L. The contribution of 30. Oborn BM, Dowdell S, Metcalfe PE, Crozier S, Mohan R, Keall PJ. Future magnetic resonance imaging to the three-dimensional treatment of Medical Physics: Real-time MRI guided Proton Therapy. Med Phys. planning of localized prostate cancer. Int J Radiat Oncol Biol Phys. 2017;44:77–90. https://doi.org/10.1002/mp.12371. 31. Walker A, Liney G, Metcalfe P, Holloway L. MRI distortion: Considerations 1999;45(4):857–65. https://doi.org/10.1016/S0360-3016(99)00288-6. 13. Dirix P, Haustermans K, Vandecaveye V. The Value of Magnetic for MRI based radiotherapy treatment planning. Australas Phys Eng Sci Resonance Imaging for Radiotherapy Planning. Semin Radiat Oncol. Med. 2014;37(1):103–13. https://doi.org/10.1007/s13246-014-0252-2. 2014;24(3):151–9. https://doi.org/10.1016/j.semradonc.2014.02.003. 32. Maspero M. MR-only Radiotherapy of prostate cancer. PhD thesis, Utrecht 14. Roach MI, Faillace-Akazawa P, Malfatti C, Holland J, Hricak H. Prostate University. 2018. volumes defined by magnetic resonance imaging and computerized 33. Zangger K, Armitage LM. Silver and gold NMR. Metal-based drugs. tomographic scans for three-dimensional conformal radiotherapy. Int J 1999;6(4-5):239–45. https://doi.org/10.1155/MBD.1999.239. Radiat Oncol Biol Phys. 1996;35(5):1011–18. https://doi.org/10.1016/ 34. Lim TY, Kudchadker RJ, Wang J, Stafford RJ, MacLellan C, Rao A, 0360-3016(96)00232-5. Ibbott GS, Frank SJ. Effect of pulse sequence parameter selection on 15. Rasch C, Barillot I, Remeijer P, Touw A, van Herk M, Lebesque JV. signal strength in positive-contrast MRI markers for MRI-based prostate Definition of the prostate in CT and MRI: a multi-observer study. Int J postimplant assessment. Med Phys. 2016;43(7):4312–22. https://doi.org/ Radiat Oncol Biol Phys. 1999;43(1):57–66. https://doi.org/10.1016/S0360- 10.1118/1.4953635. 3016(98)00351-4. 35. Jonsson JH, Garpebring A, Karlsson MG, Nyholm T. Internal fiducial 16. Villeirs GM, Vaerenbergh K, Vakaet L, Bral S, Claus F, Neve WJ, markers and susceptibility effects in MRI—simulation and measurement Verstraete KL, Meerleer GO. Interobserver Delineation Variation Using CT of spatial accuracy. Int J Radiat Oncol Biol Phys. 2012;82(5):1612–8. versus Combined CT + MRI in Intensity-Modulated Radiotherapy for https://doi.org/10.1016/j.ijrobp.2011.01.046. Prostate Cancer. Strahlenther Onkol. 2005;181(7):424–30. https://doi.org/ 36. Ghose S, Mitra J, Rivest-Hénault D, Fazlollahi A, Stanwell P, Pichler P, 10.1007/s00066-005-1383-x. Sun J, Fripp J, Greer PB, Dowling JA. MRI-alone radiation therapy 17. Parker CC, Damyanovich A, Haycocks T, Haider M, Bayley A, Catton CN. planning for prostate cancer: Automatic fiducial marker detection. Med. Magnetic resonance imaging in the radiation treatment planning of Phys. 2016;43(5):2218–28. https://doi.org/10.1118/1.4944871. localized prostate cancer using intra-prostatic fiducial markers for 37. Dinis Fernandes C, Dinh CV, Steggerda MJ, ter Beek LC, Smolic M, computed tomography co-registration. Radiot Oncol. 2003;66(2):217–24. van Buuren LD, Pos FJ, van der Heide UA. Prostate fiducial marker https://doi.org/10.1016/S0167-8140(02)00407-3. detection with the use of multi-parametric magnetic resonance imaging. 18. Huisman HJ, Fütterer JJ, van Lin ENJT, Welmers A, Scheenen TWJ, van Phys Imag Radiat Oncol. 2017;1:14–20. https://doi.org/10.1016/j.phro. Dalen Ja, Visser AG, Witjes JA, Barentsz JO. Prostate cancer: precision of 2017.02.001. integrating functional MR imaging with radiation therapy treatment by 38. Gustafsson C, Korhonen J, Persson E, Gunnlaugsson A, Nyholm T, using fiducial gold markers. Radiology. 2005;236(1):311–17. https://doi. Olsson LE. Registration free automatic identification of gold fiducial org/10.1148/radiol.2361040560. markers in MRI target delineation images for prostate radiotherapy. Med 19. Jonsson J. H, Brynolfsson P, Garpebring A, Karlsson M, Söderström K, Phys. 2017;44(11):5563–74. https://doi.org/10.1002/mp.12516. Nyholm T. Registration accuracy for MR images of the prostate using a 39. Maspero M, van den Berg CAT, Zijlstra F, Sikkes GG, de Boer HCJ, subvolume based registration protocol. Radiat Oncol. 2011;6(1):73. Meijer GJ, Kerkmeijer LGW, Viergever MA, Lagendijk JJW, Meijer GJ, https://doi.org/10.1186/1748-717X-6-73. Seevinck PR. Evaluation of an automatic MR-based gold fiducial marker Maspero et al. Radiation Oncology (2018) 13:105 Page 12 of 12 localisation method for MR-only prostate radiotherapy. Phys Med Biol. 2017;62(20):7981–8002. https://doi.org/10.1088/1361-6560/aa875f. 40. Hong CG, Yoon BI, Choe H-S, Ha U-S, Sohn DW, Cho Y-H. The Prevalence and Characteristic Differences in Prostatic Calcification between Health Promotion Center and Urology Department Outpatients. Kor J Urol. 2012;53(5):330. https://doi.org/10.4111/kju.2012.53.5.330. 41. Ung NM, Wee L. Fiducial registration error as a statistical process control metric in image-guidance radiotherapy with fiducial markers. Phys Med Biol. 2011;56(23):7473–85. https://doi.org/10.1088/0031-9155/56/23/009. 42. Tyagi N, Fontenla S, Zelefsky M, Chong-Ton M, Ostergren K, Shah N, Warner L, Kadbi M, Mechalakos J, Hunt M. Clinical workflow for mr-only simulation and planning in prostate. Radiat Oncol. 2017;12(1):119. https:// doi.org/10.1186/s13014-017-0854-4. 43. Jones M, Dobson A, O’Brian S. A graphical method for assessing agreement with the mean between multiple observers using continuous measures. Int J Epid. 2011;40(5):1308–13. https://doi.org/10.1093/ije/ dyr109. 44. Deegan T, Owen R, Holt T, Fielding A, Biggs J, Parfitt M, Coates A, Roberts L. Assessment of cone beam CT registration for prostate radiation therapy: Fiducial marker and soft tissue methods. J Med Imag Radiat Oncol. 2015;59(1):91–8. https://doi.org/10.1111/1754-9485.12197. 45. Lips IM, Dehnad H, van Gils CH, Boeken Kruger AE, van der Heide UA, van Vulpen M. High-dose intensity-modulated radiotherapy for prostate cancer using daily fiducial marker-based position verification: acute and late toxicity in 331 patients. Radiat Oncol. 2008;3:15. https://doi.org/10. 1186/1748-717X-3-15. 46. Ullman KL, Ning H, Susil R, Ayele A, Jocelyn L, Havelos J, Guion P, Xie H, Li G, Arora BC, Cannon A, Miller RW, Norman C C, Camphausen K, Ménard C. Intra- and inter-radiation therapist reproducibility of daily isocenter verification using prostatic fiducial markers. Radiat Oncol. 2006;1(1):2. https://doi.org/10.1186/1748-717X-1-2. 47. Deegan T, Owen R, Holt T, Roberts L, Biggs J, McCarthy A, Parfitt M, Fielding A. Interobserver variability of radiation therapists aligning to fiducial markers for prostate radiation therapy. J Med Imaging Radiat Oncol. 2013;57(4):519–23. https://doi.org/10.1111/1754-9485.12055. 48. van der Heide UA, Kotte ANTJ, Dehnad H, Hofman P, Lagendijk JJW, van Vulpen M. Analysis of fiducial marker-based position verification in the external beam radiotherapy of patients with prostate cancer. Radiother Oncol. 2007;82(1):38–45. https://doi.org/10.1016/j.radonc.2006.11.002. 49. Suh JH, Gardner JM, Kee KH, Shen S, Ayala AG, Ro JY. Calcifications in prostate and ejaculatory system: a study on 298 consecutive whole mount sections of prostate from radical prostatectomy or cystoprostatectomy specimens. Ann Diagn Pathol. 2008;12(3):165–70. https://doi.org/10.1016/j.anndiagpath.2007.07.001. 50. De Roover R, Crijns W, Poels K, Peeters R, Draulans C, Haustermans K, Depuydt T. Characterization of a novel liquid fiducial marker for multi-modal image guidance in stereotactic body radiotherapy of prostate cancer. Med Phys. 2018. https://doi.org/10.1002/mp.12860. 51. Bittner N, Butler WM, Reed JL, Murray BC, Kurko BS, Wallner KE, Merrick GS. Electromagnetic tracking of intrafraction prostate displacement among patients externally immobilized in the prone position. Int J Radiat Oncol Biol Phys. 2009;75(3):328. https://doi.org/10. 1016/j.ijrobp.2009.07.752. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Radiation Oncology Springer Journals

Evaluation of gold fiducial marker manual localisation for magnetic resonance-only prostate radiotherapy

Free
12 pages
Loading next page...
 
/lp/springer_journal/evaluation-of-gold-fiducial-marker-manual-localisation-for-magnetic-WkhMW40Hcm
Publisher
BioMed Central
Copyright
Copyright © 2018 by The Author(s)
Subject
Biomedicine; Cancer Research; Oncology; Radiotherapy; Imaging / Radiology
eISSN
1748-717X
D.O.I.
10.1186/s13014-018-1029-7
Publisher site
See Article on Publisher Site

Abstract

Background: The use of intraprostatic gold fiducial markers (FMs) ensures highly accurate and precise image-guided radiation therapy for patients diagnosed with prostate cancer thanks to the ease of localising FMs on photon-based imaging, like Computed Tomography (CT) images. Recently, Magnetic Resonance (MR)-only radiotherapy has been proposed to simplify the workflow and reduce possible systematic uncertainties. A critical, determining factor in the accuracy of such an MR-only simulation will be accurate FM localisation using solely MR images. Purpose: The aim of this study is to evaluate the performances of manual MR-based FM localisation within a clinical environment. Methods: We designed a study in which 5 clinically involved radiation therapy technicians (RTTs) independently localised the gold FMs implanted in 16 prostate cancer patients in two scenarios: employing a single MR sequence or a combination of sequences. Inter-observer precision and accuracy were assessed for the two scenarios for localisation in terms of 95% limit of agreement on single FMs (LoA)/ centre of mass (LoA ) and inter-marker CM distances (IDs), respectively. Results: The number of precisely located FMs (LoA<2 mm) increased from 38/48 to 45/48 FMs when localisation was performed using multiple sequences instead of single one. When performing localisation on multiple sequences, imprecise localisation of the FMs (3/48 FMs) occurred for 1/3 implanted FMs in three different patients. In terms of precision, we obtained LoA within 0.25 mm in all directions over the precisely located FMs. In terms of accuracy, IDs CM difference of manual MR-based localisation versus CT-based localisation was on average (±1STD)0.6±0.6 mm. Conclusions: For both the investigated scenarios, the results indicate that when FM classification was correct, the precision and accuracy are high and comparable to CT-based FM localisation. We found that use of multiple sequences led to better localisation performances compared with the use of single sequence. However, we observed that, due to the presence of calcification and motion, the risk of mislocated patient positioning is still too high to allow the sole use of manual FM localisation. Finally, strategies to possibly overcome the current challenges were proposed. Keywords: Magnetic resonance imaging, Radiotherapy treatment planning, MR-only treatment planning, Pre-treatment positioning, Fiducial marker localization, Manual detection, Accuracy, Precision *Correspondence: m.maspero@umcutrecht.nl; matteo.maspero.it@gmail.com Universitair Medisch Centrum Utrecht, Heidelberglaan 100, 3508 GA Utrecht, The Netherlands © 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. Maspero et al. Radiation Oncology (2018) 13:105 Page 2 of 12 Background use of MR images as reference for position verification The use of intraprostatic fiducial markers (FMs) ensures purposes [32]. highly accurate and precise image-guided radiation ther- Up to now, research on such an MR-only workflow has apy (IGRT) for patients diagnosed with prostate can- mainly focused upon generation of so-called synthetic- cer [1]. Specifically, it has been shown that markers CT images to allow dose calculation based on MR image enable for a safe reduction of PTV margin [2–5]. To information alone. Less attention has been paid to the accurately position patients based on the target loca- issue of MR-based FM localisation, which is a major deter- tion, a set of three non-co-linear markers centred in mining factor in obtaining accurate radiation treatment of the prostate is the minimum requirement allowing tri- prostate cancer. angulation and measurement of the prostate position in In MRI, FMs are depicted as signal voids in magnitude different planes [6]. Most markers are made of inert images since they do not produce nuclear magnetic res- metals (gold and titanium, for example). For prostate onance signal [33]. The appearance of FM voids varies cancer, gold FMs are generally employed and their use according to imaging parameters [34]and theFMorien- allows accurate prostate localisation [7]onphoton-based tation with respect to the magnetic field [35]. Up to now, images thanks to their high density, which increases radio to minimise manual interaction, automated MR-based FM opacity [8]. Markers have usually a cylindrical shape, a localisation methods have been proposed [36–39]. These diameter ranging between 0.5 and 1.5 mm and length methods are promising, resulting in acceptable accuracy between 2 and 10 mm; they remain in the patient per- and relatively high detection rates ranging from 84% to manently [7]. To ensure geometrically accurate IGRT 96%. However, we can not rely on the fact that all FMs will treatments, FMs are localised on computed tomography be automatically localised, leaving to manual observers (CT) during treatment simulation and they are localised the burden to correct for missed detections. These missed on kV/MV imaging before patient irradiation to verify detections can derive from misclassifying blood clots or and eventually match the pre-treatment patient position calcification as FMs since they all appear as signal voids in with the planned position [9, 10]. The localisation is MR images [7, 40]. Therefore, in addition to the require- performed based on a distinct image contrast the FMs ment of achieving a high localisation accuracy, the risk of induce in the CT as well as kV/MV images [11]. The misclassification of signal voids (i.e. false positives) should distinctness of the contrast ensures that false classifica- be very low as this could result in systematic errors in tion of FMs, e.g. due the presence of calcifications, is patient positioning [41]. unlikely. In current clinical practice, MR-based manual FM local- Recently, magnetic resonance imaging (MRI) and its isation is performed by radiation therapy technicians superior soft tissue contrast with respect to photon-based (RTTs) for registering MR to CT images [17, 41]. In such imaging [12, 13] enabled more accurate delineation of a setting, the presence of CT images greatly aids the RTT the prostate [14–16]. To exploit the advantages offered to discern whether signal voids in MR images can be by this imaging modality, the use of MRI in radiother- classified as FM, calcifications or blood clots. apy planning is rapidly expanding [6]. Nowadays, the While this MR-based FM localisation for CT/MR treatment simulation is generally based on CT and MRI simulation is clinically accepted, no previous study has images. Before treatment target delineation, CT and MR investigated the reliability of a solely MR-based FM images are rigidly registered based on the location of localisation by manual observers, which is the expected FMs on both image modalities [17, 18]. The accuracy of scenario within MR-only radiotherapy. the registration is generally considered as being within The aim of this study is to evaluate the performance of 1mm[19]. manual FM localisation during the planning of prostate More recently, MRI-based radiotherapy - also called cancer patients’ external beam radiotherapy treatments. “MR-only” radiotherapy - has been proposed [20–22] Furthermore, since in a clinical environment multiple to reduce systematic spatial uncertainties introduced sequences are usually available, we aim at investigat- when registering CT and MRI images [23]. More- ing whether the use of multiple sequences may impact over, an MR-only workflow would reduce costs of themanualFMlocalisation. We conductedastudyto the treatment and patient exposure to ionising radi- assess the inter-observer precision and accuracy of the ation [24]. Additionally, MR-only treatment planning MR-based FM localisation among RTTs. Furthermore, is particularly desirable in the context of MR-guided we evaluated in our patients’ group the occurrence of photon [25–27] and eventually proton [28–30] radio- misclassification of FMs. therapy. On the other hand, the introduction of an MR-only radiotherapy pathway raises a series of chal- Methods lenges as enabling MR-based dose calculations [22], This study is divided into three parts. First, we selected dealing with distortions in MR images [31], and the patients and acquired CT and MR images (Patient Maspero et al. Radiation Oncology (2018) 13:105 Page 3 of 12 preparation and selection section). Second, we per- MR-based fiducial marker localisation formed a multi-observer manual MR-based localisation Among the acquired MR images, we tested manual FM (MR-based fiducial marker localisation section). Finally, localisation on the three following sequences, for which we evaluated the precision and accuracy of the manual FM imaging parameters are reported in Table 1: localisation and investigated whether the observer agree- ment may lead to a precise and accurate patient alignment 1. a 3D Cartesian balanced steady-state free precession (Statistical analysis section). (bSSFP) sequence with spectral attenuated inversion recovery to obtain fat suppression and highlight Patient preparation and selection prostate boundaries. The images acquired with this The study was performed on patients with prostate carci- sequence were used by the physician to perform noma diagnosis who underwent radiotherapy planning at prostate delineations. The vendor’s name for this the University Medical Center Utrecht (The Netherlands) sequence was “3D balanced turbo field echo”. between September and October 2015. The study has 2. a 3D Cartesian T1-weighted dual radio frequency been conducted in accordance with regulations from the spoiled gradient-recalled echo (SPGR) sequence. The local ethical committee. SPGR sequence was acquired right after the bSSFP For position verification purposes, each patient received and used by the physician to distinguish bleedings three intraprostatic cylindrical gold FMs (HA2 Medizin- from the primary lesion. The vendor’s name for this technik GmbH, Germany) measuring 1 mm (diameter) by sequence was “3D T1 fast field echo”. 5 mm (length). The FMs were transperineally implanted 3. a 3D Cartesian dual gradient-recalled echo (GRE) under ultrasound guidance by a physician prior to the sequence. This sequence was acquired at the end of imaging session using two 18-gauge needles placed with the examination to have an independent sequence the aid of a template. for FM localisation. The field of view (FOV) was Patient positioning at CT scan (Brilliance CT Big Bore, reduced to the sole target reducing acquisition time Philips Medical Systems, Cleveland, Ohio, USA) was con- as well as making this sequence less prone to motion ducted simulating the treatment, i.e. using a flat table, artefacts. The vendor’s name for this sequence was knee wedges, positioning arms on the chest and tattooing “3D fast field echo”. the patient with the aid of laser alignment. Patient setup at 3T MR scan (Ingenia Omega HP, The sequences were selected as best candidates Philips Healthcare, Best, The Netherlands) was performed for FM identification after inspection of previously acquired patient images. The rationale underlying the using a knee wedge, but without a flat table top, with- out positioning arms on the chest and without laser- choice was the high contrast between FM location aided positioning. Patients were scanned using anterior and surrounding tissues and the spatial high resolution and posterior phased array coils (dS Torso and Posterior coils, 28 channels, Philips Healthcare, Best, The Nether- Table 1 Imaging parameters of the sequences used for gold FM lands). To avoid compression of patients’ anatomy, two manual localisation: the second column provides the details for in-house-built coil bridges supported the anterior coil. the balanced steady-state free precession (bSSFP) sequence, the The location of FMs from CT images was obtained as third column for the radio frequency spoiled gradient-recalled previously described in [39]. No rectum or bladder prepa- echo (SPGR) sequence acquired right after the bSSFP and the ration protocol was applied before imaging sessions, but fourth column for the gradient-recalled echo (GRE) sequence the patients were asked to empty their rectum in case rec- acquired at the end of the examination tal filling was noticed being larger than 5 cm during the Imaging parameters bSSFP SPGR GRE imaging session. TE /(TE )/TR [ms] 1.98/3.96 1.4/2.7/4.4 1.4/2.7/4.6 1 2 Criteria for selecting the subjects were: patients had Flip Angle [°] 40 10 10 gold FMs implanted prior to the imaging sessions, patients a 3 underwent CT and MRI on the same day acquiring FOV [mm ] 250x250x90 467x467x300 449x449x90 three specific MR sequences (see next section for fur- a Acquisition Matrix 252x234x90 312x314x200 376x376x75 ther details) and were without metallic implants. CT scans Reconstruction Matrix 512x512x90 320x320x200 400x400x75 were performed with the following imaging parameters: a 3 Reconstructed Voxel [mm ] 0.5x0.5x1.0 1.5x1.5x1.5 1.1x1.1x1.2 120 kV, exposure time = 923 ms, tube current between Bandwidth [Hz/voxel] 945 1078 1142 121 and 183 mA, in-plane matrix = 512x512 pixels, and Readout direction AP AP AP 3 mm slice thickness. The resolution was variable depend- ing on the field of view (FOV) used. The typical size of Acquisition time 4 min 29 s 4 min 1 s 2 min 10 s the FOV was 500x500x300 mm , which corresponds to an expressed in terms of anterior-posterior, right-left and superior-inferior directions in-plane resolution of 0.98x0.98 mm . The term FOV refers to field of view, AP to anterior-posterior Maspero et al. Radiation Oncology (2018) 13:105 Page 4 of 12 (voxel size <1.2x1.2x1.5 mm ). Furthermore, they were were identified, the FM located with the lowest reliabil- expected to preserve geometrical accuracy thanks to the ity was excluded from the statistical analysis. To generate 3D acquisition and large bandwidth (>900 Hz/pixel); sim- a consistent FM marker labelling among observers, FMs ilar sequences were used also by other institutions for FM were numbered (from 1 to 3) according to the position localisation [37, 42]. Five RTTs independently performed of their centre along the superior-inferior direction. In manual localisation of FMs by identifying the top and bot- case FMswerelocated in thesametransverseplane,the tom of the markers on magnitude images of the bSSFP left-right direction was used for labelling. To keep the con- sequence. The RTTs were requested to follow standard- sistency of the labelling among the observers, the labelling ised instructions regarding zoom and window/level of the was manually checked and, when necessary, corrected. images. The provided instructions are available as part The analysis was performed in Matlab (R2015a, the Math- of the Additional file 1. The RTTs involved in the study Works Inc., Natick, Massachusetts, United States). had varying experience as technicians: 11, 6.5, 5, 5 and 10 Inter-observer agreement or spatial precision years. They all had experience with position verification. Single FM locations For each observer, the number of Four RTTs (all the observers except the first observer), detected FMs was reported. An agreement position was had experience with image registration between MR and defined as the mean position among all the five observers. CT images of: 0, 2, 2, 2.5 and 2.5 years, respectively. They To assess the precision among the five observers, FM loca- were new to MR-only FM localisation since this procedure tions were compared calculating a 95% limit of agreement was initiated with this study. The observers were asked (= 1.96 times the standard deviation (STD) [43]) of the dis- to report for which patients the FM localisation was per- tance to the mean position (LoA) in the three directions ceived as being problematic and they got the freedom to (X = anterior-posterior, Y = right-left and Z = superior- provide three or four candidates. Also, the RTTs indicated inferior). For comparison with [44], a threshold for clinical which FM was the most difficult to distinguish. The FM acceptability was set to 95% LoA ≤ 2 mm. Bar plots centre was calculated as the mean position between the providing a visual assessment of the inter-observer vari- manually identified top and bottom positions. Note that ability were created reporting also a more stringent LoA the observed length of the FM may not correspond to threshold of 1 mm. An investigation of the CT and MR the nominal length of the FM (5mm). The term “appar- images was performed on an individual patient basis to ent” location was used to refer to the location of top and investigate the causes underlying imprecise localisation bottom as identified by the observers. Note that apparent of FMs; schematic representations of the inter-observer location of the top and bottom of each FM coincided with localisation were also examined. the centre of a voxel, thus, the calculated centre of the FMs may be located with resolution higher than a single voxel. Centre of mass locations To verify the impact on patient To investigate whether the use of multiple sequences alignment, the location of the centre of mass (CM) among impacts the FM localisation, the RTTs performed the all the FMs was calculated for each patient and observer. FM localisation again using multiple MR images after For each patient, the 95% limit of agreement (= 1.96 localisation using images of the sole bSSFP sequence. In times the standard deviation (STD) [43]) with the aver- particular, in this second localisation the RTTs employed age position of the CMs (LoA ) was calculated among CM images of the bSSFP, the second echo of the SPGR and all the observers in the three directions [44]. To assess the GRE sequences. Note that the location of the mark- agreement of the CM position among the observers a ers on the bSSFP was not made available when repeating threshold of LoA < 2mmwas used forcomparison CM the localisation. In the case of inter-scan FM motion, the with [44]. To verify clinical CM agreement, the threshold RTTs were instructed to consider the position of the FMs LoA < 1 mm was employed. Bar plots providing a CM on the bSSFP as a reference. visual assessment of the inter-observer agreement vari- For completeness, the following metrics were also ability were created. Note that patient alignment is gen- recorded: the apparent length as characterised in each erally performed on the centre of mass location [41]; sequenceand by thedifferent observers, thetimerequired therefore, this is considered as the final metric to assess by each observer to complete the FM localisation and the inter-observer precision. number of FMs for which the localisation was perceived as being problematic. These results are reported as part of Intra-observer agreement the Additional file 2. To evaluate whether, over all the observers, a statistically significant variation of FM location occurred between Statistical analysis localisation using only the bSSFP sequence and the combi- The analysis was performed on the FM centres as located nation of bSSFP, SPGR and GRE sequences, we performed using both a “single” sequence (bSSFP) and “multiple” Wilcoxon rank-sum test at the confidence level of 95% in MRI sequences (bSSFP, SPGR and GRE). In case four FMs the three directions on LoA and LoA . CM Maspero et al. Radiation Oncology (2018) 13:105 Page 5 of 12 Spatial accuracy x 5 observers x 2 sequence scenarios) performed over all The difference of inter-marker distances (IDs) between the observers. The FMs were implanted at least one week the FMs located in CT and MRI were calculated for the prior to imaging. precisely located (LoA < 2 mm) markers using single During the pre-planning imaging session, MRI scans and multiple sequences. For each observer and over all were performed within maximum 70 min (mean time the observers, the absolute difference between the ID of = 45 min, minimum time of 20 min and IQR = 34- MRI and CT were calculated as in [38] and characterised 50 min) after the CT scans. For all the patients, the in terms of mean, median, standard deviation (STD) and bSSFP and SPGR sequences were acquired one after each range ([minimum, maximum]). other with a maximum time difference of 5 min, while the GRE sequence was acquired at least 15 min after Results the SPGR sequence. Figure 1 shows a zoomed axial slice Seventeen consecutive prostate patients (61.4-81.9 years, of CT, bSSFP, SPGR and GRE images for a patient in mean age = 68.7years,medianage = 68.3 years, inter which all observers agreed on the locations of the FMs quartile range (IQR) = 66.1-70.8 years) were considered (P1, top) and a patient in which one FM was challenging forinclusion in thestudy.All thepatientswerestagedas (P9, bottom) T1c-3b, Gleason score ≥ 6 and one of the patients (P14) had a hip implant and was excluded from the analysis (the Single FM locations localisation and images for this patient are presented in All the observers detected three FMs for all the patients, the Additional file 3 in Fig. 3)). Within the patient popula- except one observer (Obs1) who detected four FMs for tion, the average prostate volume during imaging sessions two patients (P4 and P17) when using the bSSFP sequence. was 56.8 ml (range = 32.1-117.3 ml, median volume = When also the SPGR and GRE sequences were employed, 54.8 ml, IQR = 42.9-70.9 ml) and body mass index was all the RTTs localised three FMs. Figure 2 provides a 2 2 on average 26.4 kg/m (range = 19.9-30.7 kg/m ,IQR = schematic representation of the centres of the FMs as 24.9-28.7 kg/m ). No patient received adjuvant hormonal localised by all the observers for patients P1 and P9 using therapy. multiple sequences. The agreement position is as well Patients underwent intensity-modulated radiotherapy, shown. As example, taking into consideration the FM with using5beamsof10MV, with aprescribeddoseof77Gy the largest spread for patient P1 (FM 1), the LoA were 0.69, to theentireprostatein35fractions (2.2 Gy perfraction). 0.57, 0.84 mm in X (anterior-posterior), Y (right-left) and Other clinical prescriptions are specified in [45]. Each of Z (superior-inferior) directions, respectively. the patients had three FMs implanted leading to a total The bar plots of LoA for all the patients over the of 48 FMs and 580 FM localisations (16 patients x 3 FMs five RTTs is shown in Fig. 3.The LoAwas foundto Fig. 1 Zoom of an axial slice of CT (left), bSSFP (centre-left), SPGR (centre-right) and GRE (right) images for patients P1 (top) and P9 (bottom) before image registration. The axes X and Y indicate the anterior-posterior and right-left directions. The intensity of CT image is in Hounsfield Units (HU), while of MR images is normalised to the maximum over the whole dataset. Note the presence of calcifications for the patient P9; they are visible as high intensity on CT and signal void on MR images Maspero et al. Radiation Oncology (2018) 13:105 Page 6 of 12 Fig. 2 Schematic representation of the centres of the FMs as localised by all the observers for patient P1 (top) and P9 (bottom) using multiple sequences (bSSFP, SPGR and GRE). The labelling of the FM is indicated by the marker: ◦,x,+and • for FM having number 1, 2, 3 and for agreement position, respectively. For patient P1 (top), the LoA of FM1 over the five RTTs was 0.27, 0.31 and 0.38 mm in X, Y and Z, respectively, which was CM considered clinically acceptable; for patient P9 (bottom), the LoA for FM2 over the five observers was 11.17, 0.99 and 13.70 mm in X, Y and Z, CM respectively, which was considered clinically unacceptable be higher than 2 mm in one of the three directions was found to be > 2 mm for more than one FM per for 10/48 and 3/48 FMs when the observers located on patient, while when localisation was performed on mul- a single (bSSFP) and multiple (bSSFP, SPGR and GRE) tiple sequences LoA was > 2 mm only for one FM per sequences, respectively. This resulted in an increased patient. Focusing on the scenario with the largest agree- agreement (45/48) when the observers located on multi- ment (localisation performed using multiple sequences), ple sequences with respect to a single sequence (38/48). localisation of maximum one FM was found to be impre- Over all the three directions, the Wilcoxon rank-sum test cise for 3/16 patients: P4, P6 and P9. Excluding these at 95% confidence interval resulted in significantly differ- 3 FMs (considering, therefore, 45/48 FMs), the average ent LoA when comparing FM location obtained with one (±1STD)LoA was0.19 ±0.15, 0.18 ±0.12 and 0.30 or multiple sequences. In particular, as shown in Fig. 3 ±0.31 mm in anterior-posterior, right-left and superior- with the use of multiple sequences for two patients (P7 inferior directions, respectively. After the investigation of and P17) the LoA decreased below 2 mm. When local- the images acquired for the patients resulting in an impre- isingusing asinglesequenceoverall thepatients, LoA cise FM localisation, we observed that patients P4 (Fig. 1 Maspero et al. Radiation Oncology (2018) 13:105 Page 7 of 12 Fig. 3 The 95% limit of agreement (LoA) calculated, for each patient, over the five observers for the single fiducial marker (FM) in the three directions, where X = anterior-posterior (top), Y = right-left (center) and Z = superior-inferior (bottom). On the left is shown the FM localisation as performed using the bSSFP sequence only, while on the right the FM localisation as performed using multiple sequences. The dotted and dashed lines represent the LoA of 2 mm, while the dotted lines represent the LoA of 1 mm. Note that patient P14 had hip implant and the results are here presented but were excluded in the statistical analysis in Additional file 3) and P9 (Figs. 1 and 2 bottom) were bSSFP impacting reliability of the localisation for this FM. characterised by the presence of large (>2mmindiame- Considering the results from a different perspective, for ter) intra-prostatic calcifications. In both cases, 1/5 RTTs the total 240 (16 patients x 5 observers x 3 FMs) single (Obs1 for P4 and Obs3 for P9) localised one of the FMs observer localisations using multiple sequences, 2 times far away from the other four observers. Figure 2 shows calcifications were marked as FMs by one of the RTTs that a misclassification occurred for patient P9 when con- and no agreement could be found for one FM among all sidering the FM2 and Obs3. The same occurred for FM2 the five RTTs. This would result in misclassification for and Obs1 for patient P4 (Fig. 1 in Additional file 3 bot- 7 out of the 240 single observer localisations, or 7/80 (16 tom). After observing the location of the misclassified patients x 5 observers) single observer localisations of the FMsasreportedbythe twoobservers in theMRand CT CM in the case the outliers cannot be eliminated. images, we found that the FMs were located in corre- spondence of calcifications. For one patient (P6, as shown Center of mass locations in Additional file 3 in Fig. 2), one of the FMs was not visible Figure 4 presents the bar plot of the 95% LoA for CM on bSFFP but appeared on SPGR and GRE; we hypothe- all the patients over the five RTTs. The LoA was CM sised that motion reduced the visibility of the FM on the found to be >1 mm in one of the three directions (X, Fig. 4 The 95% limit of agreement of the centre of mass (LoA ) calculated, for each patient, over the five observers for a single fiducial marker (FM) CM in the three directions, where X = anterior-posterior (blue), Y = right-left (red) and Z = superior-inferior (green). On the left is shown the localisation of the CM as performed using the bSSFP sequence only, while on the right the localisation of the CM as performed using multiple sequences. The dotted and dashed lines represent the LoA of 2 mm, while the dotted lines represent the LoA of 1 mm. Note that patient P14 had hip implant and the results are here presented but were excluded in the statistical analysis Maspero et al. Radiation Oncology (2018) 13:105 Page 8 of 12 Y or Z) for 5/16 and 3/16 patients when the observers Note that this fundamentally differs from localisation of located on a single (bSSFP) and multiple (bSSFP, SPGR FMsonMRimagesinthe currentCT-MR simulation and GRE) sequences, respectively. Over all the three workflow for registration purposes as the CT images directions, the Wilcoxon rank-sum test at 95% confi- can be used to minimise misclassification on the MR dence interval resulted in significantly different LoA when images. In this sense, this study was conducted to verify comparing FM locations obtained with one or multiple whether an MR-only simulation could facilitate a robust sequences. Excluding the imprecisely located CMs, the positioning workflow comparable to current CT-based average (±1STD)LoA when localisation was per- positioning in all the cases. CM formed with single sequence was 0.10 ±0.05, 0.10 ±0.06 In this study, the use of multiple sequences led to pre- and 0.19 ±0.13 mm in anterior-posterior, right-left and cise localisation (LoA< 2 mm) in more patients and of superior-inferior directions, respectively; the average (±1 more FMs (13/16 patients and 45/48 FMs) than locali- STD) LoA when localisation was performed with mul- sation with a single sequence only (11/16 patients and CM tiple sequences was 0.11 ±0.06, 0.13 ±0.09 and 0.23 ±0.18 38/48 FMs). For both scenarios, the precision calculated mm in anterior-posterior, right-left and superior-inferior as the average of LoA in all the directions on the pre- CM directions, respectively. In all the directions, the average cise localised FMs was within 0.25 mm. The results are LoA is <0.25 mm. in good agreement with others [18, 44, 46, 47]. Huisman CM et al. [18] obtained a precision of 0.5 mm in the centroid Spatial accuracy of the prostate on a cohort of 21 patients when assess- Table 2 shows the mean, median, STD and range of the ing registration of CT and MR images. Ullman et al. [46] absolute difference in the ID of the precisely located reported a mean inter-observer variability of 0.9±0.6 mm FMs (LoA <2 mm) using single and multiple sequences. when performing registration on photon-based portal Among all the observers, the average ID difference is images. Deegan et al. [47] reported that inter-observer slightly lower (0.5±0.6 mm) when locating with mul- LoA on the applied registration, which is comparable to tiple sequences with respect to with a single sequence the LoA , was in the range of about ±2 mm. Litera- CM (0.7±0.6 mm). ture reporting single FM localisation precision has not been found. Discussion In general, in our study, when FMs were precisely The precision and accuracy of manual localisation of localised, they were also accurately localised. In particu- intraprostatic gold FMs using solely MR images was eval- lar, we found an inter-observer accuracy of 0.7 mm with the single sequence and of 0.6 mm with the multiple uated in the context of an MRI-only simulation workflow. sequences. These results are slightly more accurate than what presented when comparing a human observatory to Table 2 The mean, median, standard deviation (STD) and range automatic FM localisation by Gustafsson et al. [38]and ([min, max]) of the absolute difference in the inter-marker in line with the accuracy previously considered acceptable distances (IDs) of the precisely located FMs between CT and MRI for the CM localisation performed with photon-based for all the single observers and for all the five observers imaging (0.6 mm) [48]. Sequence Observer Mean Median STD Range However, for a single FM in 3/16 cases, precise local- isation was not achieved. That implies that a correct Single 1 0.8 0.6 0.7 [0.1, 3.1] positioning in these patients can not be guaranteed. Based 2 0.6 0.5 0.5 [0.0, 2.5] on the thorough investigation of the images of these spe- 3 0.7 0.6 0.5 [0.0, 2.1] cific patients, we concluded that the following two causes 4 0.7 0.6 0.6 [0.1, 2.9] may have led to imprecise FM localisation: (1) presence of 5 0.7 0.5 0.6 [0.1, 2.5] calcifications miscalssified as FM and (2) motion during All 0.7 0.6 0.6 [0.0, 3.1] the bSSFP sequence. (1) Previous studies reported the presence of calcifications in 40 to 88% of prostate cancer Multiple 1 0.7 0.4 0.6 [0.0, 2.7] patients [7, 38, 40, 49]. In our study, for 2/16 patients the 2 0.6 0.4 0.6 [0.0, 3.0] presence of calcifications led to misclassified FM localisa- 3 0.7 0.5 0.7 [0.0, 2.5] tion for 1/5 RTT. Interestingly, the observers seemed to be aware of the difficulties and they reported that the local- 4 0.7 0.5 0.6 [0.0, 2.8] isation procedure for such patients was problematic (see 5 0.7 0.6 0.5 [0.0, 2.5] Fig. 1 in the Additional file 2). (2) Motion as a possible All 0.6 0.5 0.6 [0.0, 3.0] cause of hampered accuracy of FM localisation has already The results were calculated excluding 10/48 and 3/48 FMs for the localisation been reported in the literature for the bSSFP sequence performed on a single and multiple sequences, respectively. All the values are [42]. The readout of this sequence was 3D leading to expressed in mm Maspero et al. Radiation Oncology (2018) 13:105 Page 9 of 12 typical acquisition times of 2–3 min, and thus motion visualisation and also manual localisation performance. 2) blurring is likely to occur. Among the available MRI sequences, only the images of To obtain accurate localisation for all the patient cases, one echo of the gradient-echo sequences have been taken we believe that redundancy should be added in the locali- into consideration in this study. It may occur that acquir- sation procedure to lower the risk of FM misclassification. ing with different image parameters or MR sequences In this sense, we foresee the following as possible ways to may result in more favourable manual FM localisation increase the redundancy: performance. For example, recently, the use of multi-echo images showed promising results, thanks to the increasing multiple observers localisation. Whenever an RTT size of a signal void when increasing the echo time [38]; would have low confidence in the FM localisation, an Future studies could investigate whether MR sequence independent observer could perform localisation and optimisation or the use of other sequences may be more assess a posteriori the initially found position. In this suitable for FM localisation, verifying accuracy and preci- scenario, the experience of the RTT may influence sion performance. the outcome. Further investigations are necessary to From a general perspective, in our study, five RTTs evaluate whether such scenario will lead to accurate were involved, making the findings representative of localisation in all the cases. a realistic situation. As the observers were not famil- implantation of a fourth marker. The use of a fourth iar with MR-only FM localisation, it may be expected FM could be easily performed without increasing the that better results may be obtained by training the patient discomfort: the fourth FM could be observers for this specific context. In this sense, it collinearly placed with the third FM avoiding a new may be interesting to verify, in a future study, the needle insertion. In case of FM misclasssification, the influence of clinical experience on manual localisation RTTs may explicitly exclude one of the FM when performance. correcting patient set-up, remaining with a sufficient Comparing our study to previous research, a limita- number of FMs to enable the procedure. On the tion of the presented cohort is its size, although, no other other hand, with four FMs several permutations of 3 research has been presented to assess manual FM locali- FMs could be considered and the RTTs would need sation with such details and reporting localisation perfor- consistently choose the FMs between imaging mances within a realistic clinical environment. Recently, modalities to obtain identical set-up corrections. Gustafsson et al. [38] presented results of the accu- resorting to automatic localisation. Given the racy of manual FM localisation and a larger cohort promising result obtained with automatic gold FM (44 patients). Unfortunately, the precision has not been localisation methods [36–39], resorting to a reported. combination of manual and automated MR-based Recently, the use of MR-visible fiducial markers have FM localisation methods may ensure safe MR-based been proposed offering new possibilities for MR-based simulation of patient position. marker localisation [34, 50]. In addition, FM localisation In our institution further investigation is ongoing to may also be based on mechanisms other than imaging. verify that using automatic localisation [39]isaviable For example, it has been shown that transponders can approach including also the insertion of a fourth FM. be safely implanted ensuring real-time prostate localisa- Alternatively to the redundancy options above proposed, tion [51]. Both these approaches may be adopted in an another centre [42] reported that using kV radiogra- MR-only workflow offering an alternative to gold FM phy after FM implantation provided independent images localisation. that facilitated MR-based FM localisation. Similarly to In the perspective of MR-only Radiotherapy, and con- this approach, we could also speculate about designing sidering the case of gold FM, the use of multiple sequences a workflow that foresees referring the patients to CT would enable manual marker localisation for precise and in case of dubious manual FM localisation at the MR accurate simulation of prostate cancer patients’ position scan. Performing a low dose CT for all the patients prior to irradiation in almost all the cases. Nevertheless, for the sole purpose of FM localisation could be believing that an MR-only simulation should facilitate a another possibility. robust positioning workflow, we think the risk of mislo- Strategies to possibly solve FM misclassification other cated patient positioning is still too high and that addi- than adding redundancy may involve 1) further MR tional redundancy is essential to enable a safe clinical sequence optimisation and 2) employing different MR practice. sequences. 1) Further MR sequence optimisation could, for example, be employed to diminish the susceptibility to Conclusion motion by reducing the acquisition time of the employed We studied inter-observer precision and accuracy of man- sequences. In addition, sequence optimisation may impact ual gold FM localisation for MR-only prostate cancer Maspero et al. Radiation Oncology (2018) 13:105 Page 10 of 12 external beam therapy simulation over five RTTs for two observers (bottom) are shown in Figs. 1 and 2, respectively. For completeness, we report also the CT ad MRI images along with the scenarios: employing a single MRI sequence (bSSFP) or schematic representation of the centres of the FM for patient P14 in Fig. 3. a combination of multiple sequences (bSSFP, SPGR and Note that this patient was not considered during the analysis since had a GRE). The use of multiple sequences (bSSFP, SPGR and hip implant. (PDF 1823 kb) GRE) led to better localisation performances compared with the use of a single sequence (bSSFP). For both the Abbreviations scenarios, the results indicate that when FM classification AP: Anterior-posterior; bSSFP: Balanced steady-state free precession; CT: Computed tomography; CM: Centre of mass; FM: Fiducial marker; FOV: Field of was correct, the precision and accuracy are high and com- view; GRE: Gradient-recalled echo; HU: Hounsfield units; IGRT: Image-guided parable to CT-based FM localisation. However, the risk radiotherapy; IQR: Inter quartile range; kV: Kilo Volt; LoA: Limit of agreement; of mislocated patient positioning due to FM misclassifi- LoA : Limit of agreement of the centre of mass; MV: Mega volt; MR(I): CM Magnetic resonance (imaging); Obs: Observer; P: Patient; PTV: Planning target cation is still too high to allow the sole use of manual volume; RTT: Radiation therapy technician; SPGR: Spoiled gradient-recalled FM localisation. For future work, we hypothesise that fur- echo; STD: Standard deviation; TE: Echo time; TR: Repetition time ther increasing redundancy by increasing the number of Acknowledgements FM per patient and by setting up a system to rely on We are grateful to Nicole Vissers, Joske Boudewijn and Tiny Vlig (UMC Utrecht, multiple observations or automatic localisation is nec- The Netherlands) for performing the manual localisation and their kind essary to increase the detection rate and enable clinical collaboration. We would like to thank Gert J Meijer (UMC Utrecht, The Netherlands) for discussion about the design of the study and Max A Viergever introduction. (UMC Utrecht, The Netherlands) and Jan J W Lagendijk (UMC Utrecht, The Netherlands) for providing general support to the research. Additional files Funding The research is funded by ZonMw IMDI Programme, project number: 1040030. The project is co-funded by Philips Healthcare. Additional file 1: Instructions provided to the clinical observers. As part of the supplementary material is possible to download a repository Availability of data and materials (InstructionPackage.zip) containing the instructions provided to the RTTs The datasets analysed during the current study are not publicly available due before performing the manual FM localisation. In particular, the repository to internal policy of the Medical Ethical Commission pertaining to data sharing contains the following files: but are available from the corresponding author upon Medical Ethical [1.] GeneralGuidelineFMloc.pdf which presents a short description of the Commission’s approval. The datasets generated during the inter-observer procedure; localisation, e.g. the FM location and the correspondent analysis are available [2.] PracticalInstructionFMloc.pdf which describes step-by-step the at https://matteomaspero.github.io/Manual-gold-FM-localisation/. procedure; [3.] Checklist_Obs.pdf which is aimed at supporting the RTTs during the Authors’ contributions procedure in keeping track and annotate for which patient the localisation MM designed and managed the study, collected and analysed the data and was found problematic. (ZIP 194 kb) drafted/revised the manuscript. PRS participated to the study design and the revision of the manuscript. NJW, GGS and GJK participated to design the study, Additional file 2: Annotations on the FM localisation. As part of the to collect inter-observer FM localisations and revise the manuscript. HCJdB supplementary material, we report the apparent length of the FMs for each and JRNvdVvZ contributed to design the study and revise the manuscript. observer and the time spent by each observer performing the FM CATvdB participated to design the study and to revise the manuscript. All localisation over all the patients. In particular, Table 1 shows the mean, authors read and approved the final manuscript. standard deviation (STD), range [min, max] of the apparent length, expressed in mm. The weighted mean over all the observer is 7.5 ± 0.6 mm Ethics approval and consent to participate and 7.7 ± 0.7 mm for localisation using a single and multiple sequences, The study received approval of the medical ethical commission (Medisch respectively. Note that the apparent length was longer than the nominal Ethische Toetsingscommissie) and was classified under the protocol number length of the FM (5 mm). Table 2 reports the mean, STD and range of the 15-444/C approved on 29th July 2015. time needed by each observer to perform the FM localisation using single and multiple sequences. The weighted mean over all the observer is Competing interests 5.8 ± 1.4 min. Note that all the RTTs localised the FMs first using a single Peter R Seevinck declares to be a majority shareholder of MRIGuidance B.V. and then multiple sequences for all the patients. The RTTs were free to Cornelis A T van den Berg declares to be a minority shareholder of MRCode B.V. chose the order of patients and whether concluding the procedure first for each the patients using both single and multiple sequences or first for all Publisher’s Note the patients using single sequence and then repeat for all the patients Springer Nature remains neutral with regard to jurisdictional claims in using multiple modalities. Possible differences in the way the RTTs published maps and institutional affiliations. performed the procedure does not permit to understand whether the FM localisation is faster using single or multiple sequences. In addition, a Received: 27 October 2017 Accepted: 13 April 2018 histogram reporting the frequency of unreliable FM localisation, as perceived by the RTTs is shown in Fig. 1 for four out of five observers; one of the observers did not report the reliability of the localisation. The observers References reported the perceived reliability without distinction between localisation 1. Zaorsky NG, Showalter TN, Ezzell GA, Nguyen PL, Assimos DG, performed employing a single and multiple sequences. (PDF 108 kb) D’Amico AV, Gottschalk AR, Gustafson GS, Keole SR, Liauw SL, Lloyd S, Additional file 3: Single patient investigation. As a supplementary McLaughlin PW, Movsas B, Prestidge BR, Taira AV, Vapiwala N, Davis BJ. material, we report CT and MRI images for the patients P4 and P6, which ACR appropriateness criteria® external beam radiation therapy treatment were found having LoA > 2 mm in maximum one of the three FMs for planning for clinically localized prostate cancer, Part II of II. Advances localisation performed with multiple sequences. Zoom of an axial slice of Radiat Oncol. 2017. https://doi.org/10.1016/j.adro.2017.03.003. CT (top left), bSSFP (top centre-left), SPGR (top centre-right) and GRE (top 2. Wu J, Haycocks T, Alasti H, Ottewell G, Middlemiss N, Abdolell M, right) images for the patients P4, P6 before image registration as well as Warde P, Toi A, Catton C. Positioning errors and prostate motion during schematic representations of the centres of the FMs as localised by all the conformal prostate radiotherapy using on-line isocentre set-up Maspero et al. Radiation Oncology (2018) 13:105 Page 11 of 12 verification and implanted prostate markers. Radiother Oncol. 2001;61(2): 20. Fraass BA, McShan DL, Diaz RF, Ten Haken RK, Aisen A, Gebarski S, 127–33. https://doi.org/10.1016/S0167-8140(01)00452-2. Glazer G, Lichter AS. Integration of magnetic resonance imaging into 3. Schallenkamp JM, Herman MG, Kruse JJ, Pisansky TM. Prostate position radiation therapy treatment planning: i. technical considerations. Int J relative to pelvic bony anatomy based on intraprostatic gold markers and Radiat Oncol Biol Phys. 1987;13(12):1897–908. https://doi.org/10.1016/ electronic portal imaging. Int J Radiat Oncol Biol Phys. 2005;63(3):800–11. 0360-3016(87)90358-0. https://doi.org/10.1016/j.ijrobp.2005.02.022. 21. Lee YK, Bollet M, Charles-Edwards G, Flower MA, Leach MO, McNair H, 4. Beltran C, Herman MG, Davis BJ. Planning Target Margin Calculations for Moore E, Rowbottom C, Webb S. Radiotherapy treatment planning of Prostate Radiotherapy Based on Intrafraction and Interfraction Motion prostate cancer using magnetic resonance imaging alone. Radiother Oncol. Using Four Localization Methods. Int J Radiat Oncol Biol Phys. 2008;70(1): 2003;66(2):203–16. https://doi.org/10.1016/S0167-8140(02)00440-1. 289–95. https://doi.org/10.1016/j.ijrobp.2007.08.040. 22. Edmund JM, Nyholm T. A review of substitute CT generation for MRI-only 5. Greer P, Dahl K, Ebert M, Wratten C, White M, Denham J. Comparison of radiation therapy. Radiat Oncol. 2017;12(1):28. https://doi.org/10.1186/ prostate set-up accuracy and margins with off-line bony anatomy s13014-016-0747-y. corrections and online implanted fiducial-based corrections. J Med 23. Nyholm T, Nyberg M, Karlsson MG. Systematisation of spatial Imaging Radiat Oncol. 2008;52(5):511–6. https://doi.org/10.1111/j.1440- uncertainties for comparison between a MR and a CT-based radiotherapy 1673.2008.02005.x. workflow for prostate treatments. Radiat Oncol. 2009;4(1):54. https://doi. 6. Schmidt MA, Payne GS. Radiotherapy planning using MRI. Phys Md Biol. org/10.1186/1748-717X-4-54. 2015;60(22):323–61. https://doi.org/10.1088/0031-9155/60/22/R323. 24. Karlsson M, Karlsson M. G, Nyholm T, Amies C, Zackrisson B. Dedicated 7. Ng M, Brown E, Williams A, Chao M, Lawrentschuk N, Chee R. Fiducial Magnetic Resonance Imaging in the Radiotherapy Clinic. Int J Radiat markers and spacers in prostate radiotherapy: current applications. BJU Oncol Biol Phys. 2009;74(2):644–51. https://doi.org/10.1016/j.ijrobp.2009. Int. 2014;113(S2):13–20. https://doi.org/10.1111/bju.12624. 01.065. 8. Gall K. P, Verhey L. J, Wagner M. Computer-assisted positioning of 25. Raaymakers BW, Raaijmakers AJE, Kotte ANTJ, Jette D, Lagendijk JJW. radiotherapy patients using implanted radiopaque fiducials. Med. Phys. Integrating a MRI scanner with a 6 MV radiotherapy accelerator: dose 1993;20(4):1153–9. https://doi.org/10.1118/1.596969. deposition in a transverse magnetic field. Phys Med Biol. 2004;49(17): 9. Balter JM, Lam KL, Sandler HM, Littles JF, Bree RL, Ten Haken RK. 4109–18. https://doi.org/10.1088/0031-9155/49/17/019. Automated localization of the prostate at the time of treatment using 26. Dempsey J, Benoit D, Fitzsimmons J, Haghighat A, Li J, Low D, Mutic S, implanted radiopaque markers: Technical feasibility. Int J Radiat Oncol Biol Palta J, Romeijn H, Sjoden G. A device for realtime 3D image-guided Phys. 1995;33(5):1281–6. https://doi.org/10.1016/0360-3016(95)02083-7. IMRT. Int J Radiat Oncol Biol Phys. 2005;63:202. 10. Vigneault E, Pouliot J, Laverdière J, Roy J, Dorion M. Electronic portal 27. Fallone BG, Murray B, Rathee S, Stanescu T, Steciw S, Vidakovic S, imaging device detection of radioopaque markers for the evaluation of Blosser E, Tymofichuk D. First MR images obtained during megavoltage prostate position during megavoltage irradiation: A clinical study. Int J photon irradiation from a prototype integrated linac-MR system. Med Radiat Oncol Biol Phys. 1997;37(1):205–12. https://doi.org/10.1016/S0360- Phys. 2009;36(6):2084–8. https://doi.org/10.1118/1.3125662. 3016(96)00341-0. 28. Raaymakers BW, Raaijmakers AJE, Lagendijk JJW. Feasibility of MRI 11. Habermehl D, Henkner K, Ecker S, Jäkel O, Debus J, Combs SE. guided proton therapy: magnetic field dose effects. Phys Med Biol. Evaluation of different fiducial markers for image-guided radiotherapy 2008;53(20):5615–22. https://doi.org/10.1088/0031-9155/53/20/003. and particle therapy. J. Radiat. Res. 2013;54(Suppl 1):61–8. https://doi.org/ 29. Moteabbed M, Schuemann J, Paganetti H. Dosimetric feasibility of 10.1093/jrr/rrt071. real-time MRI-guided proton therapy. Med Phys. 2014;41(11):111713. 12. Debois M, Oyen R, Maes F, Verswijvel G, Gatti G, Bosmans H, Feron M, https://doi.org/10.1118/1.4897570. Bellon E, Kutcher G, Van Poppel H, Vanuytsel L. The contribution of 30. Oborn BM, Dowdell S, Metcalfe PE, Crozier S, Mohan R, Keall PJ. Future magnetic resonance imaging to the three-dimensional treatment of Medical Physics: Real-time MRI guided Proton Therapy. Med Phys. planning of localized prostate cancer. Int J Radiat Oncol Biol Phys. 2017;44:77–90. https://doi.org/10.1002/mp.12371. 31. Walker A, Liney G, Metcalfe P, Holloway L. MRI distortion: Considerations 1999;45(4):857–65. https://doi.org/10.1016/S0360-3016(99)00288-6. 13. Dirix P, Haustermans K, Vandecaveye V. The Value of Magnetic for MRI based radiotherapy treatment planning. Australas Phys Eng Sci Resonance Imaging for Radiotherapy Planning. Semin Radiat Oncol. Med. 2014;37(1):103–13. https://doi.org/10.1007/s13246-014-0252-2. 2014;24(3):151–9. https://doi.org/10.1016/j.semradonc.2014.02.003. 32. Maspero M. MR-only Radiotherapy of prostate cancer. PhD thesis, Utrecht 14. Roach MI, Faillace-Akazawa P, Malfatti C, Holland J, Hricak H. Prostate University. 2018. volumes defined by magnetic resonance imaging and computerized 33. Zangger K, Armitage LM. Silver and gold NMR. Metal-based drugs. tomographic scans for three-dimensional conformal radiotherapy. Int J 1999;6(4-5):239–45. https://doi.org/10.1155/MBD.1999.239. Radiat Oncol Biol Phys. 1996;35(5):1011–18. https://doi.org/10.1016/ 34. Lim TY, Kudchadker RJ, Wang J, Stafford RJ, MacLellan C, Rao A, 0360-3016(96)00232-5. Ibbott GS, Frank SJ. Effect of pulse sequence parameter selection on 15. Rasch C, Barillot I, Remeijer P, Touw A, van Herk M, Lebesque JV. signal strength in positive-contrast MRI markers for MRI-based prostate Definition of the prostate in CT and MRI: a multi-observer study. Int J postimplant assessment. Med Phys. 2016;43(7):4312–22. https://doi.org/ Radiat Oncol Biol Phys. 1999;43(1):57–66. https://doi.org/10.1016/S0360- 10.1118/1.4953635. 3016(98)00351-4. 35. Jonsson JH, Garpebring A, Karlsson MG, Nyholm T. Internal fiducial 16. Villeirs GM, Vaerenbergh K, Vakaet L, Bral S, Claus F, Neve WJ, markers and susceptibility effects in MRI—simulation and measurement Verstraete KL, Meerleer GO. Interobserver Delineation Variation Using CT of spatial accuracy. Int J Radiat Oncol Biol Phys. 2012;82(5):1612–8. versus Combined CT + MRI in Intensity-Modulated Radiotherapy for https://doi.org/10.1016/j.ijrobp.2011.01.046. Prostate Cancer. Strahlenther Onkol. 2005;181(7):424–30. https://doi.org/ 36. Ghose S, Mitra J, Rivest-Hénault D, Fazlollahi A, Stanwell P, Pichler P, 10.1007/s00066-005-1383-x. Sun J, Fripp J, Greer PB, Dowling JA. MRI-alone radiation therapy 17. Parker CC, Damyanovich A, Haycocks T, Haider M, Bayley A, Catton CN. planning for prostate cancer: Automatic fiducial marker detection. Med. Magnetic resonance imaging in the radiation treatment planning of Phys. 2016;43(5):2218–28. https://doi.org/10.1118/1.4944871. localized prostate cancer using intra-prostatic fiducial markers for 37. Dinis Fernandes C, Dinh CV, Steggerda MJ, ter Beek LC, Smolic M, computed tomography co-registration. Radiot Oncol. 2003;66(2):217–24. van Buuren LD, Pos FJ, van der Heide UA. Prostate fiducial marker https://doi.org/10.1016/S0167-8140(02)00407-3. detection with the use of multi-parametric magnetic resonance imaging. 18. Huisman HJ, Fütterer JJ, van Lin ENJT, Welmers A, Scheenen TWJ, van Phys Imag Radiat Oncol. 2017;1:14–20. https://doi.org/10.1016/j.phro. Dalen Ja, Visser AG, Witjes JA, Barentsz JO. Prostate cancer: precision of 2017.02.001. integrating functional MR imaging with radiation therapy treatment by 38. Gustafsson C, Korhonen J, Persson E, Gunnlaugsson A, Nyholm T, using fiducial gold markers. Radiology. 2005;236(1):311–17. https://doi. Olsson LE. Registration free automatic identification of gold fiducial org/10.1148/radiol.2361040560. markers in MRI target delineation images for prostate radiotherapy. Med 19. Jonsson J. H, Brynolfsson P, Garpebring A, Karlsson M, Söderström K, Phys. 2017;44(11):5563–74. https://doi.org/10.1002/mp.12516. Nyholm T. Registration accuracy for MR images of the prostate using a 39. Maspero M, van den Berg CAT, Zijlstra F, Sikkes GG, de Boer HCJ, subvolume based registration protocol. Radiat Oncol. 2011;6(1):73. Meijer GJ, Kerkmeijer LGW, Viergever MA, Lagendijk JJW, Meijer GJ, https://doi.org/10.1186/1748-717X-6-73. Seevinck PR. Evaluation of an automatic MR-based gold fiducial marker Maspero et al. Radiation Oncology (2018) 13:105 Page 12 of 12 localisation method for MR-only prostate radiotherapy. Phys Med Biol. 2017;62(20):7981–8002. https://doi.org/10.1088/1361-6560/aa875f. 40. Hong CG, Yoon BI, Choe H-S, Ha U-S, Sohn DW, Cho Y-H. The Prevalence and Characteristic Differences in Prostatic Calcification between Health Promotion Center and Urology Department Outpatients. Kor J Urol. 2012;53(5):330. https://doi.org/10.4111/kju.2012.53.5.330. 41. Ung NM, Wee L. Fiducial registration error as a statistical process control metric in image-guidance radiotherapy with fiducial markers. Phys Med Biol. 2011;56(23):7473–85. https://doi.org/10.1088/0031-9155/56/23/009. 42. Tyagi N, Fontenla S, Zelefsky M, Chong-Ton M, Ostergren K, Shah N, Warner L, Kadbi M, Mechalakos J, Hunt M. Clinical workflow for mr-only simulation and planning in prostate. Radiat Oncol. 2017;12(1):119. https:// doi.org/10.1186/s13014-017-0854-4. 43. Jones M, Dobson A, O’Brian S. A graphical method for assessing agreement with the mean between multiple observers using continuous measures. Int J Epid. 2011;40(5):1308–13. https://doi.org/10.1093/ije/ dyr109. 44. Deegan T, Owen R, Holt T, Fielding A, Biggs J, Parfitt M, Coates A, Roberts L. Assessment of cone beam CT registration for prostate radiation therapy: Fiducial marker and soft tissue methods. J Med Imag Radiat Oncol. 2015;59(1):91–8. https://doi.org/10.1111/1754-9485.12197. 45. Lips IM, Dehnad H, van Gils CH, Boeken Kruger AE, van der Heide UA, van Vulpen M. High-dose intensity-modulated radiotherapy for prostate cancer using daily fiducial marker-based position verification: acute and late toxicity in 331 patients. Radiat Oncol. 2008;3:15. https://doi.org/10. 1186/1748-717X-3-15. 46. Ullman KL, Ning H, Susil R, Ayele A, Jocelyn L, Havelos J, Guion P, Xie H, Li G, Arora BC, Cannon A, Miller RW, Norman C C, Camphausen K, Ménard C. Intra- and inter-radiation therapist reproducibility of daily isocenter verification using prostatic fiducial markers. Radiat Oncol. 2006;1(1):2. https://doi.org/10.1186/1748-717X-1-2. 47. Deegan T, Owen R, Holt T, Roberts L, Biggs J, McCarthy A, Parfitt M, Fielding A. Interobserver variability of radiation therapists aligning to fiducial markers for prostate radiation therapy. J Med Imaging Radiat Oncol. 2013;57(4):519–23. https://doi.org/10.1111/1754-9485.12055. 48. van der Heide UA, Kotte ANTJ, Dehnad H, Hofman P, Lagendijk JJW, van Vulpen M. Analysis of fiducial marker-based position verification in the external beam radiotherapy of patients with prostate cancer. Radiother Oncol. 2007;82(1):38–45. https://doi.org/10.1016/j.radonc.2006.11.002. 49. Suh JH, Gardner JM, Kee KH, Shen S, Ayala AG, Ro JY. Calcifications in prostate and ejaculatory system: a study on 298 consecutive whole mount sections of prostate from radical prostatectomy or cystoprostatectomy specimens. Ann Diagn Pathol. 2008;12(3):165–70. https://doi.org/10.1016/j.anndiagpath.2007.07.001. 50. De Roover R, Crijns W, Poels K, Peeters R, Draulans C, Haustermans K, Depuydt T. Characterization of a novel liquid fiducial marker for multi-modal image guidance in stereotactic body radiotherapy of prostate cancer. Med Phys. 2018. https://doi.org/10.1002/mp.12860. 51. Bittner N, Butler WM, Reed JL, Murray BC, Kurko BS, Wallner KE, Merrick GS. Electromagnetic tracking of intrafraction prostate displacement among patients externally immobilized in the prone position. Int J Radiat Oncol Biol Phys. 2009;75(3):328. https://doi.org/10. 1016/j.ijrobp.2009.07.752.

Journal

Radiation OncologySpringer Journals

Published: Jun 5, 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