Validation of the Vectra XT three-dimensional imaging system for measuring breast volume and symmetry following oncological reconstruction

Validation of the Vectra XT three-dimensional imaging system for measuring breast volume and... Purpose Three-dimensional surface imaging (3D-SI) of the breasts enables the measurement of breast volume and shape symmetry. If these measurements were sufficiently accurate and repeatable, they could be used in planning oncological breast surgery and as an objective measure of aesthetic outcome. The aim of this study was to validate the measurements of breast volume and symmetry provided by the Vectra XT imaging system. Methods To validate measurements, breast phantom models of true volume between 100 and 1000 cm were constructed and varying amounts removed to mimic breast tissue ‘resections’. The volumes of the phantoms were measured using 3D-SI by two observers and compared to a gold standard. For intra-observer repeatability and inter-observer reproducibility in vivo, 16 patients who had undergone oncological breast surgery had breast volume and symmetry measured three times by two observers. Results A mean relative difference of 2.17 and 2.28% for observer 1 and 2 respectively was seen in the phantom measure- ments compared to the gold standard (n = 45, Bland Altman agreement). Intra-observer variation over ten repeated meas- urements demonstrated mean coefficients of variation (CV) of 0.58 and 0.49%, respectively. The inter-observer variation demonstrated a mean relative difference of 0.11% between the two observers. In patients, intra-observer variation over three repeated volume measurements for each observer was 3.9 and 3.8% (mean CV); the mean relative difference between observ - ers was 5.78%. For three repeated shape symmetry measurements using RMS projection difference between the two breasts, the intra-observer variations were 8 and 14% (mean CV), the mean relative difference between observers was 0.43 mm for average symmetry values that ranged from about 3.5 to 15.5 mm. Conclusion This first validation of breast volume and shape symmetry measurements using the Vectra XT 3D-SI system suggests that these measurements have the potential to assist in pre-operative planning and also as a measure of aesthetic outcome. Keywords Breast cancer · Breast volume · Aesthetic outcome · Breast conservation · Breast surgery * Jennifer E. Rusby Cancer Research UK Cancer Imaging Centre, Institute jennifer.rusby@rmh.nhs.uk of Cancer Research and Royal Marsden NHS Foundation Trust, Sutton, UK Department of Breast Surgery, The Royal Marsden NHS Westmead Breast Cancer Institute, Westmead, NSW 2145, Foundation Trust, Downs Road, Sutton, Surrey SM2 5PT, Australia UK Department of Statistics, The Royal Marsden NHS Foundation Trust, Sutton, UK Joint Department of Physics and Cancer Research UK Cancer Imaging Centre, Institute of Cancer Research and Royal Marsden NHS Foundation Trust, Sutton, UK Vol.:(0123456789) 1 3 392 Breast Cancer Research and Treatment (2018) 171:391–398 Introduction Three-dimensional surface imaging (3D-SI), first described by Burke and Beard [1, 2] in 1967 to analyse facial structures, is now also used in the aesthetic breast surgery setting, mainly as a marketing tool. Patients can visualize three-dimensional (3D) images of themselves and their plastic surgeon can manipulate these images to simulate the expected appearances after breast augmen- tation or reduction. Clinicians and researchers can also exploit 3D-SI to measure breast volume and symmetry in pre-operative planning, to investigate volume reten- tion after lipofilling and use it as a measure of cosmetic Fig. 1 Image of breast phantom made of plasticine taken using the outcome after surgery [3]. Moreover, the ability to meas- Vectra XT camera. The observer has identified a region of inter - est around the perimeter of the phantom that includes an area 2  cm ure the shape symmetry, which reports not only on the beyond the boundary of the model volume but also on the surface shape, is advantageous [4]. By reflecting the mirror image of one of the patient’s breasts onto the other, the distance between the two breast (5), 400 (5), 500 (9) and 1000 (10) cm . The gold standard surfaces may be calculated over the entire surface of the volume of each phantom was measured initially and after breast from the superimposed images. However, 3D-SI has every resection using the Archimedes’ principle of water not hitherto been exploited in a breast cancer population. displacement. The phantoms were placed on a wooden In these patients, where body mass index is higher [5] board held 1.5 metres from, and parallel to, the Vectra compared to the slim build of the aesthetic breast augmen- XT. 3D-SI of the phantoms initially and after every resec- tation population [6], 3D-SI may not give accurate vol- tion was carried out using a standard protocol available ume measurements as it requires an assumption about the in the Vectra Analysis Module (VAM) software (Canfield depth and curvature of the underlying chest wall. Thus, the Sci, New Jersey, USA). By identifying a region of inter- validity of the volume and symmetry measurements in this est (ROI) including an area bounded by a 2-cm perimeter population remains unknown and its use as a clinical tool drawn on the 3D image beyond the boundary of the phan- is not established. The aim of this study, therefore, was tom, the posterior wall of the phantom was interpolated to validate breast volume measurement using the Vectra from the ROI data and the volume enclosed by the breast XT (Canfield Sci, New Jersey, USA), a popular 3D imag- phantom and its estimated posterior wall was calculated. ing system used worldwide, and to investigate the repro- Two observers independently measured the volume for ducibility of measuring breast volume and symmetry in a each phantom 10 times, from which the mean volume and breast cancer population who had undergone oncological coefficient of variation (CV) were calculated. breast surgery. The volume measurement accuracy was obtained by comparing the mean volumes for each phantom measured by 3D-SI to the gold standard. Agreement between the two methods was determined using Bland Altman plots [7] dis- Methods playing the relative difference between the mean of the ten estimates and the gold standard volume ((mean 3D-SI vol- Phantom model evaluation ume − gold standard volume)/gold standard × 100) as a function of the gold standard volume, and ‘limits of agree- The breast phantoms were made of plasticine (NCP ment’ (overall mean relative difference +/− 1.96SD of the Newclay Products, Devon, UK). Six phantoms were con- mean relative difference). structed, each of a different volume (Fig.  1), incorporating To assess intra-observer variation, the coefficient of vari- a wide range of breast volumes. ‘Resections’ were sequen- ation (CV = SD/mean) for the 10 measurements of each of tially excised from each phantom and after each resection phantom was calculated and then the mean of the CV was the phantom was re-imaged. The resections were used to calculated as an indication of the variation in repeated meas- represent typical excision biopsies, wide local excisions urements for all of the phantoms. and therapeutic mammoplasties. In total, 45 phantoms Inter-observer agreement was measured using a Bland were imaged with the following starting volumes (num- Altman plot showing the relative difference between the ber of resections in parentheses): 100 (5), 200 (5), 300 means of ten estimates for each observer ((mean volume 1 3 Breast Cancer Research and Treatment (2018) 171:391–398 393 observer 1  −  mean volume observer 2)/(mean of mean 1. Vertically down the midline from one square below the volumes for observer 1 and 2) × 100) as a function of the suprasternal notch to one square below the most medial mean of the two observers’ mean volume measurements, aspect of the infra-mammary fold (IMF). and ‘limits of agreement’ (overall mean relative difference 2. Then moving laterally at a level one square below the +/− 1.96 SD of the mean relative difference). IMF to the anterior axillary line following parallel to the curve of the IMF or ptosis of breast (whichever is more In vivo patient evaluations caudal). 3. Then moving superiorly along the anterior axillary line Following Regional Ethics Committee approval (REC num- towards the axilla. ber 14/YH/0175) 16 patients were recruited to the study. 4. Then moving medially to define the superior border to Informed consent was obtained from all individual partici- meet the midline one square below the clavicle. pants included in the study. Additional informed consent was obtained from the participant whose images are included in The volume measurements were undertaken by two this article. Participants were positioned 1.5 m away from observers imaging each patient three times. The left and the Vectra XT. The participant’s face was excluded from all right breasts were treated independently. images. Participants were imaged with their hands on their To assess intra-observer variation, the mean CV for all hips at the end of normal inspiration. breasts was calculated. Bland Altman plots were used to measure inter-observer variation and agreement as described Measurement of volume for the phantom studies. After attempting to measure volume using simple place- Measurement of shape symmetry ment of landmarks and the in-built software, and secondly by simply marking an ROI by eye, we concluded that there Similarly, a protocol was developed to measure symmetry. was too much variation and this had an unacceptable impact For each 3D image: on reproducibility. Therefore to calculate the breast volume using the VAM software (Canfield Sci, New Jersey, USA), a 1. The torso was rotated to anterior-posterior view (AP) specific protocol was developed (Fig.  2). A grid was placed view (Fig. 3a). on the image of the breast in the x, y, z planes, with each grid 2. The gridlines were place onto the image so that the y cube having a side of 2 cm. The y axis was positioned at the axis (x = 0) bisected the torso. The image was cropped midline, with the upper border of the grid at the suprasternal so that only the breast area was visible. (Fig. 3b). notch. A region of interest (ROI) was identified around the 3. The torso was viewed cranio-caudally and the shoul- perimeter of the breast as follows: ders/clavicles aligned along the line of z = 0 in case the patient was not standing parallel to the Vectra XT cam- eras at the time of imaging (Fig. 3c). 4. The image was rotated 90° laterally and cropped either side at the level of the anterior axillary line (Fig. 3d, e). 5. The image was returned to the AP position. One-half of the torso was selected from the y axis laterally using a lasso tool provided within the VAM software (Fig. 3f). 6. The whole image was copied and reflected with the y axis as the line of symmetry, inverting the z-axis data so that in both the original and the reflected images the z position of the breast surface was positive. Using the VAM software, the distances along the z axis (skin sur- face ‘height’) between the corresponding x–y coordi- nate points on the original image and the reflected image were measured. (Fig. 3g). This measurement was termed the height difference (HD). The degree of asymmetry was then quantified using the RMS projection differ - ence (RMS-PD), as the square root of the mean of the squared HDs, which gives a positive value regardless of whether left is larger than right or vice versa. The lower the number, the more symmetrical the breasts is. Fig. 2 Method developed to measure volume using Vectra-XT 3D-SI 1 3 394 Breast Cancer Research and Treatment (2018) 171:391–398 Fig. 3 Protocol to measure breast symmetry, RMS Projec- tion difference (RMS-PD). a AP view of torso, b gridlines placed onto image and cropped, c torso in craniocaudal view, d image rotated laterally, e image cropped at level of the anterior axillary line, f one-half of torso is selected, Image is copied and reflected in x = 0 Both observers calculated the RMS-PD for each of the Repeatability and reproducibility of volume and shape three images of each patient. To assess intra-observer vari- symmetry measurements in patients ation, the mean of the CV was calculated. A Bland Altman plot was used to measure inter-observer agreement and vari- Of the 18 women invited to participate in this validation ation as above. study, 16 agreed. The other two were unable to attend in Data were entered into an excel spread sheet (Microsoft the allocated timeframe. Eleven women had undergone Corp., Redmond, Washington) and analysed using SPSS sta- mastectomy and reconstruction and 5 had undergone breast tistical software (SPSS v22; SPSS, Inc., Chicago). conservation. The average age was 53 years (SD = 9 years) 2 2 and the average BMI was 24.8 kg/m (SD = 4.9 kg/m ). Results Volume measurements Phantom model evaluations The mean CV over three repeated volume measurements Accuracy was 3.9% and 3.8% for observer 1 and 2, respectively. No significant correlation was observed between CV and The mean relative differences for observers 1 and 2 com- breast volume (correlation coefficients 0.025 and − 0.17, pared with the gold standard were 2.17 and 2.28% with lim- and p values 0.891 and 0.366, respectively, for observer its of agreement of − 0.42 to 4.76% and − 0.33 to 4.95%, 1 and 2), thus justifying the averaging of CV over breast respectively (Fig. 4a, b). volume. The Bland Altman plot (Fig. 6) demonstrated a mean rela- Intra-observer variability tive difference for 3D-SI volume between the two observers of − 5.78% with limits of agreement of − 17.76 to 6.2% for The mean CVs were 0.58 and 0.49% for observers 1 and 2, patients’ breast volume measurements obtained by averaging respectively. No significant correlation was seen between over three estimates. CV and volume (correlation coefficients 0.026 and 0.273, and p values 0.867 and 0.069, respectively, for observer 1 and 2), thus justifying the averaging of CV over phantoms Shape symmetry measurements of different volumes. The mean CV of repeated volume measurements was 8.0 and Inter-observer agreement and variability 14.0% for observer 1 and 2, respectively. The Bland Altman plot (Fig.  7) demonstrated that the The Bland Altman plot that compared one observer’s vol- mean difference between the observers’ estimates of symme- ume measurements with the other’s (Fig. 5) demonstrated try was 0.43 mm, for average symmetry values that ranged a mean relative difference for 3D-SI volume measurements from about 3.5 mm to about 15.5 mm, with upper limit of between the two observers of 0.11% and limits of agreement agreement of 4.01  mm and lower limit of agreement of of − 0.63 to 0.84%. − 3.15 mm. 1 3 Breast Cancer Research and Treatment (2018) 171:391–398 395 Fig. 4 a Bland Altman plot of a 1 agreement between gold stand- ard and mean Vectra XT for Observer 1 for phantom breast model volumes, b Bland Alt- man plot of agreement between -1 gold standard and mean Vectra XT for observer 2 for phantom -2 breast model volumes -3 -4 -5 -6 0 200 400 600 80010001200 Relative mean difference Upper limit of agreement Lower limit of agreement Gold standard volume (cm ) -1 -2 -3 -4 -5 -6 0 200 400600 800 1000 1200 Relative mean difference Upper limit of agreement Lower limit of agreement Gold standard volume (cm ) Fig. 5 Bland Altman plot of 1.5 agreement between observers 1 and 2 for phantom breast model 1.0 volumes 0.5 0.0 -0.5 -1.0 -1.5 0 200 400600 80010001200 Average of the observers 1 and 2 (cm ) Relative mean difference Upper limits of agreement Lower limits of agreement 1 3 Relative difference of observers 1 and Relativedifference of mean 3D-SI volume for Relative difference of mean 3D-SI volume for 2 measurement of breast phantom observer 2 and gold standard (%) observer 1 and gold standard (%) volume using 3D-SI (%) 396 Breast Cancer Research and Treatment (2018) 171:391–398 Fig. 6 Bland Altman plot of agreement between observ- ers 1 and 2 for patients’ breast volumes -10 -20 -30 -40 -50 0 200 400600 800 10001200 Average of observer 1 and 2 3D-SI volume (cm ) Relative mean difference Upper limit of agreement Lower limit of agreement Fig. 7 Bland Altman plot of 7 agreement between observers 1 and 2 for breast symmetry -1 -2 -3 -4 0246 810121416 Average of observer 1 and 2 RMS projection difference (mm) Mean difference Upper limit Lower limit 300  cm , and for larger volumes there may even have Discussion been a slight overestimation. These trends were consistent across the two observers. The intra-observer variation was We have demonstrated the ability to assess the volume and found to be low (mean CV 0.58% over ten estimates for shape symmetry of breasts in a breast cancer population one observer and 0.49% for the other). The two observers with prior oncological surgery and shown a high degree of strongly agreed; inter-observer agreement obtained from reproducibility between observers. Invasive breast cancer the Bland Altman plot (Fig.  5) was on average within affects approximately 55,000 women a year in the United 0.11% of the volume measured i.e. there was no significant Kingdom [8] and at least 80% have some form of surgery bias of one observer relative to the other compared with [9] that necessitates the best possible aesthetic outcome the residual SD of 0.38% seen in Fig. 5. because it influences psychological recovery and qual- A similar study of a 3D-SI system designed and made ity of life [10–12]. The ability to objectively measure the in-house by researchers at the University of Glasgow volume and symmetry of the breasts has the potential to was validated using a phantom of the breast [13]. In their profoundly influence surgical planning and thus aesthetic study, the authors found that on average, relative to the outcome. gold standard, their 3D-SI volume measurements overes- As was shown in Fig.  4a, b, the study demonstrated timated the volumes by 5%, whereas in this study the bias that the Vectra XT is able to measure the volume of a was slightly smaller at about 2% underestimation. Another simple breast-shaped object with an average accuracy study using phantoms to validate the Minolta Vivid 910 of about 2.2% underestimation of the true volume (one (Konica Minolta Co Ltd, Osaka, Japan) 3D-SI system observer − 2.17%, the other − 2.28%). Variation in accu- was undertaken by Kovacs et al. [14] who used ‘dummy racy over the volume range studied suggested that most models’ of the breasts consisting of two mannequins. The of this underestimation occurred at volumes smaller than authors found that there was a lower variance (CV) for 1 3 Difference between observer 1 and 2 Relative difference of observer 1 and 2 RMS projection difference calculation measurement of breast volume using 3D-SI(%) (mm) Breast Cancer Research and Treatment (2018) 171:391–398 397 repeatability of breast volume measurements performed The RMS-PD or similar methodology has been used by an experienced observer compared to those with less as a measure of shape symmetry in other studies [19–23]. experience (lowest CV 1.28%, highest 2.5%). They related In our study, population RMS-PD ranged between 3 and this to the ability of the observers to accurately mark the 16 mm (see Fig. 7). Another study [24] found that the aver- breast borders on the thoracic wall as per a defined and age symmetry score measured using RMS-PD had a range standardised protocol. In our experiment using the plasti- of 1.7–12.8 mm in a cohort of 87 women without a history cine phantoms, the boundaries of the phantom were eas- of breast surgery. It was significantly higher in patients with ily identifiable, which explains why our intra- and inter- a higher BMI, cup size, and chest wall circumference. As observer variabilities were low. expected, the RMS-PD values in our study were higher since The second part of the present study aimed to test the they had all undergone surgery. intra-observer and inter-observer variation of human breast An objective symmetry score has the potential to be used volume and symmetry measurements. The intra-observer for measurement of aesthetic outcome after surgery. It is variation in the human breasts was on average 3.9% for one important, however, to understand what repeatability and observer and 3.8% for the other, which are much larger than reproducibility are necessary for a symmetry score to be the values obtained using the phantoms (around 0.5%). This useful in this way. A minimum requirement is that the score is likely to be due to both the complexity of defining the is significantly different for different subjective assessments region of interest in human breasts compared to the simple of aesthetic outcome. One study [20] compared objective outline of the phantoms, and the fact that only three esti- symmetry scores of this type with subjective scores for mates were made in each human breast whereas ten esti- symmetry in forty-four patients after immediate unilateral mates were obtained of each phantom’s volume. We chose breast reconstruction with extended latissimus dorsi flap. to use three measurements of the breast volume as in clini- There was a highly significant correlation (p < 0.0001, cor- cal practice it would not be practical to measure the breasts relation coefficient = 0.62) between the two scores. We also more than several times due to time constraints. The mean recently published a study demonstrating that RMS-PD sym- relative difference for volume measurements between the metry significantly agrees (Kruskal Wallis test, p < 0.001) two observers was 5.78%, which was smaller than the resid- with panel assessment of aesthetic outcome in women who ual SD of 6.1% on this Bland Altman plot but probably sig- have undergone breast conserving therapy [25]. In that study, nificant in terms of a bias of one observer compared with the patients were grouped according to panel assessment of other. Cochrane et al. [15] identified that excision of greater the aesthetic outcome after surgery (poor/fair/good/excel- than 10% of the breast volume led to a reduction in patient lent). The median RMS-PD scores for each group differed satisfaction. In addition, Sigurdson et al. [16] showed that by approximately 2 mm from those in the next sequential the minimum volume difference detectable by the human group. Therefore, the repeatability (intra-observer highest eye was 50 cm by subjective judgment. Therefore, we assert CV < 14%, highest symmetry score in patient 15.5  mm, that a relative mean difference of 5.78%, which was main- therefore worst SD 2.2  mm) and reproducibility (inter- tained for breasts volumes up to 1000 cm and thus equal to observer mean relative difference < 0.5 mm) measured here about 58 cm at the largest volume, is adequate. should be sufficient for the method to have clinical value. Mailey et al. [17] conducted a study in 22 patients under- This study has demonstrated satisfactory repeatability going breast augmentation whereby two observers measured and reproducibility of breast volume and symmetry meas- the breast volumes before and after augmentation to investi- urements using the Vectra XT three-dimensional surface gate the accuracy of measuring breast and implant volume. imaging system. However, these data provide preliminary They used the Portrait 3D surgical simulation platform (Axis results in a small cohort of patients. Before widespread clini- Three, Boston, Massachusetts, USA). The authors found no cal use, further validation should be undertaken by inde- significant difference between their two observers’ estima- pendent breast surgical research groups in a larger cohort of tions of breast volumes pre- or post-operatively, and high patients with a wide variety of body habitus, and who have reliability (intra-class correlation coefficient, ICC all > 0.94). undergone a variety of breast surgery. In another study, Losken et al. [18] measured the volume of 19 breasts of patients using the 3dMD system (3dMD LLC, Atlanta, Georgia, USA) before they underwent mastectomy Conclusion and used water displacement of the mastectomy specimen as the gold standard. They concluded that intra-observer reli- We are the first to validate breast volume and symmetry ability was excellent with an ICC of 0.975 and inter-observer measurement using the popular Vectra XT 3D-SI, which is reliability was also excellent with an inter-class correlation essential before widespread use. For volume measurement, coefficient of 0.968. However, no quantitative measurements the intra-observer variation was low (CV < 4%), and mean of repeatability or reproducibility were provided. difference between the two observers was acceptable at 1 3 398 Breast Cancer Research and Treatment (2018) 171:391–398 6. Singh N, Picha GJ, Murphy DK (2016) Natrelle silicone breast about 5.8%. For symmetry of the breasts, the intra-observer implant follow-up study: demographics, lifestyle, and surgical variation was low (CV < 14%) and mean inter-observer dif- characteristics of more than 50,000 augmentation subjects. Plast ference was 0.43 mm in a study population for which the Reconstr Surg 137(1):70–81 range of symmetry values was 3–16 mm. This repeatability 7. Bland JM, Altman DG (1986) Statistical methods for assessing agreement between two methods of clinical measurement. Lancet and reproducibility would be adequate for clinical use in 1(8476):307–410 pre-operative planning of breast surgery and as measures 8. UK, C.R. http://www.cance rrese archu k.org/healt h-profe ssion al/ of aesthetic outcome after breast cancer surgery. However, cance r-stati stics /stati stics -by-cance r-type/breas t-cance r. 2015 further validation is required before widespread clinical use. 9. Cheung S, Greenway N, Lagord C, Williams L, Kearins O, Law- rence G. h t t p : // w w w . c a n c e r s c r e e n i ng . n h s. u k / b r e a s t s c r e e n / a l l- breas t-cance r-repor t.pdf. 2006. Acknowledgements This work was possible due to a grant for a one 10. Al-Ghazal SK, Fallowfield L, Blamey RW (1999) Does cosmetic year fellowship from the Royal College of Surgeons of England. We outcome from treatment of primary breast cancer influence psy - also thank the Royal Marsden Hospital/ Institute of Cancer Research chosocial morbidity? Eur J Surg Oncol 25(6):571–573 NIHR Biomedical Research Centre for their support with this project. 11. Waljee JF et al (2008) Effect of esthetic outcome after breast- conserving surgery on psychosocial functioning and quality of Funding Royal College Surgeons of England one year Research Fel- life. J Clin Oncol 26(20):3331–3337 lowship (awarded to ROC). The Royal Marsden/Institute of Cancer 12. Heil J et al (2010) Aesthetic and functional results after breast Research is an NIHR Biomedical Research Centre. This support is conserving surgery as correlates of quality of life measured by a acknowledged. German version of the Breast Cancer Treatment Outcome Scale (BCTOS). Breast 19(6):470–474 Compliance with ethical standards 13. Henseler H et al (2011) Investigation into accuracy and repro- ducibility of a 3D breast imaging system using multiple stereo cameras. J Plast Reconstr Aesthet Surg 64(5):577–582 Conflict of interest There are no conflicts of interest. 14. Kovacs L et al (2006) New aspects of breast volume measure- ment using 3-dimensional surface imaging. Ann Plast Surg Ethical approval All procedures performed in studies involving human 57(6):602–610 participants were in accordance with the ethical standards of the insti- 15. Cochrane RA et al (2003) Cosmesis and satisfaction after breast- tutional and/or national research committee and with the 1964 Helsinki conserving surgery correlates with the percentage of breast vol- Declaration and its later amendments or comparable ethical standard. ume excised. Br J Surg 90(12):1505–1509 The protocol was reviewed and passed by Yorkshire and the Humber— 16. Sigurdson LJ, Kirkland SA (2006) Breast volume determination South Yorkshire. REC reference 14YH0175. in breast hypertrophy: an accurate method using two anthropo- morphic measurements. Plast Reconstr Surg 118(2):313–320 Informed consent Informed consent was obtained from all individual 17. Mailey B et al (2013) Clinical accuracy and reproducibility of participants included in the study. Additional informed consent was Portrait 3D Surgical Simulation Platform in breast augmentation. obtained from the individual participant for whom identifying informa- Aesthet Surg J 33(1):84–92 tion is included in this article. 18. Losken A et al (2005) Validating three-dimensional imaging of the breast. Ann Plast Surg 54(5):471–476 Open Access This article is distributed under the terms of the Crea- 19. Onesti MG et al (2004) Breast asymmetry: a new vision of this tive Commons Attribution 4.0 International License (http://creat iveco malformation. Acta Chir Plast 46(1):8–11 mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- 20. Henseler H et  al (2013) Subjective versus objective assess- tion, and reproduction in any medium, provided you give appropriate ment of breast reconstruction. J Plast Reconstr Aesthet Surg credit to the original author(s) and the source, provide a link to the 66(5):634–639 Creative Commons license, and indicate if changes were made. 21. Henseler H et al (2012) Objective evaluation of the latissimus dorsi flap for breast reconstruction using three-dimensional imag- ing. J Plast Reconstr Aesthet Surg 65(9):1209–1215 22. Moyer HR et al (2008) Three-dimensional digital evaluation of References breast symmetry after breast conservation therapy. J Am Coll Surg 207(2):227–232 1. Burke PH et al (1983) Stereophotographic measurement of change 23. Liu C et al (2010) The role of three-dimensional scanning tech- in facial soft tissue morphology following surgery. Br J Oral Surg nique in evaluation of breast asymmetry in breast augmentation: 21(4):237–245 a 100-case study. Plast Reconstr Surg 126(6):2125–2132 2. Burke PH, Beard LF, Stereo-photogrammetry of the face. Rep 24. Losken A et al (2005) An objective evaluation of breast symmetry Congr Eur Orthod Soc, 1967: 279–293 and shape differences using 3-dimensional images. Ann Plast Surg 3. O’Connell RL et al (2015) Review of three-dimensional (3D) sur- 55(6):571–575 face imaging for oncoplastic, reconstructive and aesthetic breast 25. O’Connell RL et al (2017) The potential role of three-dimen- surgery. Breast 24(4):331–342 sional surface imaging as a tool to evaluate aesthetic outcome 4. Nahabedian MY, Galdino G (2003) Symmetrical breast recon- after Breast Conserving Therapy (BCT). Breast Cancer Res Treat struction: is there a role for three-dimensional digital photogra- 164(2):385–393 phy? Plast Reconstr Surg 112(6):1582–1590 5. O’Connell RL et al (2016) Initial experience of the BREAST- Q breast-conserving therapy module. Breast Cancer Res Treat 160(1):79–89 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Breast Cancer Research and Treatment Springer Journals

Validation of the Vectra XT three-dimensional imaging system for measuring breast volume and symmetry following oncological reconstruction

Free
8 pages

Loading next page...
 
/lp/springer_journal/validation-of-the-vectra-xt-three-dimensional-imaging-system-for-XabqbDKtQ9
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Oncology
ISSN
0167-6806
eISSN
1573-7217
D.O.I.
10.1007/s10549-018-4843-6
Publisher site
See Article on Publisher Site

Abstract

Purpose Three-dimensional surface imaging (3D-SI) of the breasts enables the measurement of breast volume and shape symmetry. If these measurements were sufficiently accurate and repeatable, they could be used in planning oncological breast surgery and as an objective measure of aesthetic outcome. The aim of this study was to validate the measurements of breast volume and symmetry provided by the Vectra XT imaging system. Methods To validate measurements, breast phantom models of true volume between 100 and 1000 cm were constructed and varying amounts removed to mimic breast tissue ‘resections’. The volumes of the phantoms were measured using 3D-SI by two observers and compared to a gold standard. For intra-observer repeatability and inter-observer reproducibility in vivo, 16 patients who had undergone oncological breast surgery had breast volume and symmetry measured three times by two observers. Results A mean relative difference of 2.17 and 2.28% for observer 1 and 2 respectively was seen in the phantom measure- ments compared to the gold standard (n = 45, Bland Altman agreement). Intra-observer variation over ten repeated meas- urements demonstrated mean coefficients of variation (CV) of 0.58 and 0.49%, respectively. The inter-observer variation demonstrated a mean relative difference of 0.11% between the two observers. In patients, intra-observer variation over three repeated volume measurements for each observer was 3.9 and 3.8% (mean CV); the mean relative difference between observ - ers was 5.78%. For three repeated shape symmetry measurements using RMS projection difference between the two breasts, the intra-observer variations were 8 and 14% (mean CV), the mean relative difference between observers was 0.43 mm for average symmetry values that ranged from about 3.5 to 15.5 mm. Conclusion This first validation of breast volume and shape symmetry measurements using the Vectra XT 3D-SI system suggests that these measurements have the potential to assist in pre-operative planning and also as a measure of aesthetic outcome. Keywords Breast cancer · Breast volume · Aesthetic outcome · Breast conservation · Breast surgery * Jennifer E. Rusby Cancer Research UK Cancer Imaging Centre, Institute jennifer.rusby@rmh.nhs.uk of Cancer Research and Royal Marsden NHS Foundation Trust, Sutton, UK Department of Breast Surgery, The Royal Marsden NHS Westmead Breast Cancer Institute, Westmead, NSW 2145, Foundation Trust, Downs Road, Sutton, Surrey SM2 5PT, Australia UK Department of Statistics, The Royal Marsden NHS Foundation Trust, Sutton, UK Joint Department of Physics and Cancer Research UK Cancer Imaging Centre, Institute of Cancer Research and Royal Marsden NHS Foundation Trust, Sutton, UK Vol.:(0123456789) 1 3 392 Breast Cancer Research and Treatment (2018) 171:391–398 Introduction Three-dimensional surface imaging (3D-SI), first described by Burke and Beard [1, 2] in 1967 to analyse facial structures, is now also used in the aesthetic breast surgery setting, mainly as a marketing tool. Patients can visualize three-dimensional (3D) images of themselves and their plastic surgeon can manipulate these images to simulate the expected appearances after breast augmen- tation or reduction. Clinicians and researchers can also exploit 3D-SI to measure breast volume and symmetry in pre-operative planning, to investigate volume reten- tion after lipofilling and use it as a measure of cosmetic Fig. 1 Image of breast phantom made of plasticine taken using the outcome after surgery [3]. Moreover, the ability to meas- Vectra XT camera. The observer has identified a region of inter - est around the perimeter of the phantom that includes an area 2  cm ure the shape symmetry, which reports not only on the beyond the boundary of the model volume but also on the surface shape, is advantageous [4]. By reflecting the mirror image of one of the patient’s breasts onto the other, the distance between the two breast (5), 400 (5), 500 (9) and 1000 (10) cm . The gold standard surfaces may be calculated over the entire surface of the volume of each phantom was measured initially and after breast from the superimposed images. However, 3D-SI has every resection using the Archimedes’ principle of water not hitherto been exploited in a breast cancer population. displacement. The phantoms were placed on a wooden In these patients, where body mass index is higher [5] board held 1.5 metres from, and parallel to, the Vectra compared to the slim build of the aesthetic breast augmen- XT. 3D-SI of the phantoms initially and after every resec- tation population [6], 3D-SI may not give accurate vol- tion was carried out using a standard protocol available ume measurements as it requires an assumption about the in the Vectra Analysis Module (VAM) software (Canfield depth and curvature of the underlying chest wall. Thus, the Sci, New Jersey, USA). By identifying a region of inter- validity of the volume and symmetry measurements in this est (ROI) including an area bounded by a 2-cm perimeter population remains unknown and its use as a clinical tool drawn on the 3D image beyond the boundary of the phan- is not established. The aim of this study, therefore, was tom, the posterior wall of the phantom was interpolated to validate breast volume measurement using the Vectra from the ROI data and the volume enclosed by the breast XT (Canfield Sci, New Jersey, USA), a popular 3D imag- phantom and its estimated posterior wall was calculated. ing system used worldwide, and to investigate the repro- Two observers independently measured the volume for ducibility of measuring breast volume and symmetry in a each phantom 10 times, from which the mean volume and breast cancer population who had undergone oncological coefficient of variation (CV) were calculated. breast surgery. The volume measurement accuracy was obtained by comparing the mean volumes for each phantom measured by 3D-SI to the gold standard. Agreement between the two methods was determined using Bland Altman plots [7] dis- Methods playing the relative difference between the mean of the ten estimates and the gold standard volume ((mean 3D-SI vol- Phantom model evaluation ume − gold standard volume)/gold standard × 100) as a function of the gold standard volume, and ‘limits of agree- The breast phantoms were made of plasticine (NCP ment’ (overall mean relative difference +/− 1.96SD of the Newclay Products, Devon, UK). Six phantoms were con- mean relative difference). structed, each of a different volume (Fig.  1), incorporating To assess intra-observer variation, the coefficient of vari- a wide range of breast volumes. ‘Resections’ were sequen- ation (CV = SD/mean) for the 10 measurements of each of tially excised from each phantom and after each resection phantom was calculated and then the mean of the CV was the phantom was re-imaged. The resections were used to calculated as an indication of the variation in repeated meas- represent typical excision biopsies, wide local excisions urements for all of the phantoms. and therapeutic mammoplasties. In total, 45 phantoms Inter-observer agreement was measured using a Bland were imaged with the following starting volumes (num- Altman plot showing the relative difference between the ber of resections in parentheses): 100 (5), 200 (5), 300 means of ten estimates for each observer ((mean volume 1 3 Breast Cancer Research and Treatment (2018) 171:391–398 393 observer 1  −  mean volume observer 2)/(mean of mean 1. Vertically down the midline from one square below the volumes for observer 1 and 2) × 100) as a function of the suprasternal notch to one square below the most medial mean of the two observers’ mean volume measurements, aspect of the infra-mammary fold (IMF). and ‘limits of agreement’ (overall mean relative difference 2. Then moving laterally at a level one square below the +/− 1.96 SD of the mean relative difference). IMF to the anterior axillary line following parallel to the curve of the IMF or ptosis of breast (whichever is more In vivo patient evaluations caudal). 3. Then moving superiorly along the anterior axillary line Following Regional Ethics Committee approval (REC num- towards the axilla. ber 14/YH/0175) 16 patients were recruited to the study. 4. Then moving medially to define the superior border to Informed consent was obtained from all individual partici- meet the midline one square below the clavicle. pants included in the study. Additional informed consent was obtained from the participant whose images are included in The volume measurements were undertaken by two this article. Participants were positioned 1.5 m away from observers imaging each patient three times. The left and the Vectra XT. The participant’s face was excluded from all right breasts were treated independently. images. Participants were imaged with their hands on their To assess intra-observer variation, the mean CV for all hips at the end of normal inspiration. breasts was calculated. Bland Altman plots were used to measure inter-observer variation and agreement as described Measurement of volume for the phantom studies. After attempting to measure volume using simple place- Measurement of shape symmetry ment of landmarks and the in-built software, and secondly by simply marking an ROI by eye, we concluded that there Similarly, a protocol was developed to measure symmetry. was too much variation and this had an unacceptable impact For each 3D image: on reproducibility. Therefore to calculate the breast volume using the VAM software (Canfield Sci, New Jersey, USA), a 1. The torso was rotated to anterior-posterior view (AP) specific protocol was developed (Fig.  2). A grid was placed view (Fig. 3a). on the image of the breast in the x, y, z planes, with each grid 2. The gridlines were place onto the image so that the y cube having a side of 2 cm. The y axis was positioned at the axis (x = 0) bisected the torso. The image was cropped midline, with the upper border of the grid at the suprasternal so that only the breast area was visible. (Fig. 3b). notch. A region of interest (ROI) was identified around the 3. The torso was viewed cranio-caudally and the shoul- perimeter of the breast as follows: ders/clavicles aligned along the line of z = 0 in case the patient was not standing parallel to the Vectra XT cam- eras at the time of imaging (Fig. 3c). 4. The image was rotated 90° laterally and cropped either side at the level of the anterior axillary line (Fig. 3d, e). 5. The image was returned to the AP position. One-half of the torso was selected from the y axis laterally using a lasso tool provided within the VAM software (Fig. 3f). 6. The whole image was copied and reflected with the y axis as the line of symmetry, inverting the z-axis data so that in both the original and the reflected images the z position of the breast surface was positive. Using the VAM software, the distances along the z axis (skin sur- face ‘height’) between the corresponding x–y coordi- nate points on the original image and the reflected image were measured. (Fig. 3g). This measurement was termed the height difference (HD). The degree of asymmetry was then quantified using the RMS projection differ - ence (RMS-PD), as the square root of the mean of the squared HDs, which gives a positive value regardless of whether left is larger than right or vice versa. The lower the number, the more symmetrical the breasts is. Fig. 2 Method developed to measure volume using Vectra-XT 3D-SI 1 3 394 Breast Cancer Research and Treatment (2018) 171:391–398 Fig. 3 Protocol to measure breast symmetry, RMS Projec- tion difference (RMS-PD). a AP view of torso, b gridlines placed onto image and cropped, c torso in craniocaudal view, d image rotated laterally, e image cropped at level of the anterior axillary line, f one-half of torso is selected, Image is copied and reflected in x = 0 Both observers calculated the RMS-PD for each of the Repeatability and reproducibility of volume and shape three images of each patient. To assess intra-observer vari- symmetry measurements in patients ation, the mean of the CV was calculated. A Bland Altman plot was used to measure inter-observer agreement and vari- Of the 18 women invited to participate in this validation ation as above. study, 16 agreed. The other two were unable to attend in Data were entered into an excel spread sheet (Microsoft the allocated timeframe. Eleven women had undergone Corp., Redmond, Washington) and analysed using SPSS sta- mastectomy and reconstruction and 5 had undergone breast tistical software (SPSS v22; SPSS, Inc., Chicago). conservation. The average age was 53 years (SD = 9 years) 2 2 and the average BMI was 24.8 kg/m (SD = 4.9 kg/m ). Results Volume measurements Phantom model evaluations The mean CV over three repeated volume measurements Accuracy was 3.9% and 3.8% for observer 1 and 2, respectively. No significant correlation was observed between CV and The mean relative differences for observers 1 and 2 com- breast volume (correlation coefficients 0.025 and − 0.17, pared with the gold standard were 2.17 and 2.28% with lim- and p values 0.891 and 0.366, respectively, for observer its of agreement of − 0.42 to 4.76% and − 0.33 to 4.95%, 1 and 2), thus justifying the averaging of CV over breast respectively (Fig. 4a, b). volume. The Bland Altman plot (Fig. 6) demonstrated a mean rela- Intra-observer variability tive difference for 3D-SI volume between the two observers of − 5.78% with limits of agreement of − 17.76 to 6.2% for The mean CVs were 0.58 and 0.49% for observers 1 and 2, patients’ breast volume measurements obtained by averaging respectively. No significant correlation was seen between over three estimates. CV and volume (correlation coefficients 0.026 and 0.273, and p values 0.867 and 0.069, respectively, for observer 1 and 2), thus justifying the averaging of CV over phantoms Shape symmetry measurements of different volumes. The mean CV of repeated volume measurements was 8.0 and Inter-observer agreement and variability 14.0% for observer 1 and 2, respectively. The Bland Altman plot (Fig.  7) demonstrated that the The Bland Altman plot that compared one observer’s vol- mean difference between the observers’ estimates of symme- ume measurements with the other’s (Fig. 5) demonstrated try was 0.43 mm, for average symmetry values that ranged a mean relative difference for 3D-SI volume measurements from about 3.5 mm to about 15.5 mm, with upper limit of between the two observers of 0.11% and limits of agreement agreement of 4.01  mm and lower limit of agreement of of − 0.63 to 0.84%. − 3.15 mm. 1 3 Breast Cancer Research and Treatment (2018) 171:391–398 395 Fig. 4 a Bland Altman plot of a 1 agreement between gold stand- ard and mean Vectra XT for Observer 1 for phantom breast model volumes, b Bland Alt- man plot of agreement between -1 gold standard and mean Vectra XT for observer 2 for phantom -2 breast model volumes -3 -4 -5 -6 0 200 400 600 80010001200 Relative mean difference Upper limit of agreement Lower limit of agreement Gold standard volume (cm ) -1 -2 -3 -4 -5 -6 0 200 400600 800 1000 1200 Relative mean difference Upper limit of agreement Lower limit of agreement Gold standard volume (cm ) Fig. 5 Bland Altman plot of 1.5 agreement between observers 1 and 2 for phantom breast model 1.0 volumes 0.5 0.0 -0.5 -1.0 -1.5 0 200 400600 80010001200 Average of the observers 1 and 2 (cm ) Relative mean difference Upper limits of agreement Lower limits of agreement 1 3 Relative difference of observers 1 and Relativedifference of mean 3D-SI volume for Relative difference of mean 3D-SI volume for 2 measurement of breast phantom observer 2 and gold standard (%) observer 1 and gold standard (%) volume using 3D-SI (%) 396 Breast Cancer Research and Treatment (2018) 171:391–398 Fig. 6 Bland Altman plot of agreement between observ- ers 1 and 2 for patients’ breast volumes -10 -20 -30 -40 -50 0 200 400600 800 10001200 Average of observer 1 and 2 3D-SI volume (cm ) Relative mean difference Upper limit of agreement Lower limit of agreement Fig. 7 Bland Altman plot of 7 agreement between observers 1 and 2 for breast symmetry -1 -2 -3 -4 0246 810121416 Average of observer 1 and 2 RMS projection difference (mm) Mean difference Upper limit Lower limit 300  cm , and for larger volumes there may even have Discussion been a slight overestimation. These trends were consistent across the two observers. The intra-observer variation was We have demonstrated the ability to assess the volume and found to be low (mean CV 0.58% over ten estimates for shape symmetry of breasts in a breast cancer population one observer and 0.49% for the other). The two observers with prior oncological surgery and shown a high degree of strongly agreed; inter-observer agreement obtained from reproducibility between observers. Invasive breast cancer the Bland Altman plot (Fig.  5) was on average within affects approximately 55,000 women a year in the United 0.11% of the volume measured i.e. there was no significant Kingdom [8] and at least 80% have some form of surgery bias of one observer relative to the other compared with [9] that necessitates the best possible aesthetic outcome the residual SD of 0.38% seen in Fig. 5. because it influences psychological recovery and qual- A similar study of a 3D-SI system designed and made ity of life [10–12]. The ability to objectively measure the in-house by researchers at the University of Glasgow volume and symmetry of the breasts has the potential to was validated using a phantom of the breast [13]. In their profoundly influence surgical planning and thus aesthetic study, the authors found that on average, relative to the outcome. gold standard, their 3D-SI volume measurements overes- As was shown in Fig.  4a, b, the study demonstrated timated the volumes by 5%, whereas in this study the bias that the Vectra XT is able to measure the volume of a was slightly smaller at about 2% underestimation. Another simple breast-shaped object with an average accuracy study using phantoms to validate the Minolta Vivid 910 of about 2.2% underestimation of the true volume (one (Konica Minolta Co Ltd, Osaka, Japan) 3D-SI system observer − 2.17%, the other − 2.28%). Variation in accu- was undertaken by Kovacs et al. [14] who used ‘dummy racy over the volume range studied suggested that most models’ of the breasts consisting of two mannequins. The of this underestimation occurred at volumes smaller than authors found that there was a lower variance (CV) for 1 3 Difference between observer 1 and 2 Relative difference of observer 1 and 2 RMS projection difference calculation measurement of breast volume using 3D-SI(%) (mm) Breast Cancer Research and Treatment (2018) 171:391–398 397 repeatability of breast volume measurements performed The RMS-PD or similar methodology has been used by an experienced observer compared to those with less as a measure of shape symmetry in other studies [19–23]. experience (lowest CV 1.28%, highest 2.5%). They related In our study, population RMS-PD ranged between 3 and this to the ability of the observers to accurately mark the 16 mm (see Fig. 7). Another study [24] found that the aver- breast borders on the thoracic wall as per a defined and age symmetry score measured using RMS-PD had a range standardised protocol. In our experiment using the plasti- of 1.7–12.8 mm in a cohort of 87 women without a history cine phantoms, the boundaries of the phantom were eas- of breast surgery. It was significantly higher in patients with ily identifiable, which explains why our intra- and inter- a higher BMI, cup size, and chest wall circumference. As observer variabilities were low. expected, the RMS-PD values in our study were higher since The second part of the present study aimed to test the they had all undergone surgery. intra-observer and inter-observer variation of human breast An objective symmetry score has the potential to be used volume and symmetry measurements. The intra-observer for measurement of aesthetic outcome after surgery. It is variation in the human breasts was on average 3.9% for one important, however, to understand what repeatability and observer and 3.8% for the other, which are much larger than reproducibility are necessary for a symmetry score to be the values obtained using the phantoms (around 0.5%). This useful in this way. A minimum requirement is that the score is likely to be due to both the complexity of defining the is significantly different for different subjective assessments region of interest in human breasts compared to the simple of aesthetic outcome. One study [20] compared objective outline of the phantoms, and the fact that only three esti- symmetry scores of this type with subjective scores for mates were made in each human breast whereas ten esti- symmetry in forty-four patients after immediate unilateral mates were obtained of each phantom’s volume. We chose breast reconstruction with extended latissimus dorsi flap. to use three measurements of the breast volume as in clini- There was a highly significant correlation (p < 0.0001, cor- cal practice it would not be practical to measure the breasts relation coefficient = 0.62) between the two scores. We also more than several times due to time constraints. The mean recently published a study demonstrating that RMS-PD sym- relative difference for volume measurements between the metry significantly agrees (Kruskal Wallis test, p < 0.001) two observers was 5.78%, which was smaller than the resid- with panel assessment of aesthetic outcome in women who ual SD of 6.1% on this Bland Altman plot but probably sig- have undergone breast conserving therapy [25]. In that study, nificant in terms of a bias of one observer compared with the patients were grouped according to panel assessment of other. Cochrane et al. [15] identified that excision of greater the aesthetic outcome after surgery (poor/fair/good/excel- than 10% of the breast volume led to a reduction in patient lent). The median RMS-PD scores for each group differed satisfaction. In addition, Sigurdson et al. [16] showed that by approximately 2 mm from those in the next sequential the minimum volume difference detectable by the human group. Therefore, the repeatability (intra-observer highest eye was 50 cm by subjective judgment. Therefore, we assert CV < 14%, highest symmetry score in patient 15.5  mm, that a relative mean difference of 5.78%, which was main- therefore worst SD 2.2  mm) and reproducibility (inter- tained for breasts volumes up to 1000 cm and thus equal to observer mean relative difference < 0.5 mm) measured here about 58 cm at the largest volume, is adequate. should be sufficient for the method to have clinical value. Mailey et al. [17] conducted a study in 22 patients under- This study has demonstrated satisfactory repeatability going breast augmentation whereby two observers measured and reproducibility of breast volume and symmetry meas- the breast volumes before and after augmentation to investi- urements using the Vectra XT three-dimensional surface gate the accuracy of measuring breast and implant volume. imaging system. However, these data provide preliminary They used the Portrait 3D surgical simulation platform (Axis results in a small cohort of patients. Before widespread clini- Three, Boston, Massachusetts, USA). The authors found no cal use, further validation should be undertaken by inde- significant difference between their two observers’ estima- pendent breast surgical research groups in a larger cohort of tions of breast volumes pre- or post-operatively, and high patients with a wide variety of body habitus, and who have reliability (intra-class correlation coefficient, ICC all > 0.94). undergone a variety of breast surgery. In another study, Losken et al. [18] measured the volume of 19 breasts of patients using the 3dMD system (3dMD LLC, Atlanta, Georgia, USA) before they underwent mastectomy Conclusion and used water displacement of the mastectomy specimen as the gold standard. They concluded that intra-observer reli- We are the first to validate breast volume and symmetry ability was excellent with an ICC of 0.975 and inter-observer measurement using the popular Vectra XT 3D-SI, which is reliability was also excellent with an inter-class correlation essential before widespread use. For volume measurement, coefficient of 0.968. However, no quantitative measurements the intra-observer variation was low (CV < 4%), and mean of repeatability or reproducibility were provided. difference between the two observers was acceptable at 1 3 398 Breast Cancer Research and Treatment (2018) 171:391–398 6. Singh N, Picha GJ, Murphy DK (2016) Natrelle silicone breast about 5.8%. For symmetry of the breasts, the intra-observer implant follow-up study: demographics, lifestyle, and surgical variation was low (CV < 14%) and mean inter-observer dif- characteristics of more than 50,000 augmentation subjects. Plast ference was 0.43 mm in a study population for which the Reconstr Surg 137(1):70–81 range of symmetry values was 3–16 mm. This repeatability 7. Bland JM, Altman DG (1986) Statistical methods for assessing agreement between two methods of clinical measurement. Lancet and reproducibility would be adequate for clinical use in 1(8476):307–410 pre-operative planning of breast surgery and as measures 8. UK, C.R. http://www.cance rrese archu k.org/healt h-profe ssion al/ of aesthetic outcome after breast cancer surgery. However, cance r-stati stics /stati stics -by-cance r-type/breas t-cance r. 2015 further validation is required before widespread clinical use. 9. Cheung S, Greenway N, Lagord C, Williams L, Kearins O, Law- rence G. h t t p : // w w w . c a n c e r s c r e e n i ng . n h s. u k / b r e a s t s c r e e n / a l l- breas t-cance r-repor t.pdf. 2006. Acknowledgements This work was possible due to a grant for a one 10. Al-Ghazal SK, Fallowfield L, Blamey RW (1999) Does cosmetic year fellowship from the Royal College of Surgeons of England. We outcome from treatment of primary breast cancer influence psy - also thank the Royal Marsden Hospital/ Institute of Cancer Research chosocial morbidity? Eur J Surg Oncol 25(6):571–573 NIHR Biomedical Research Centre for their support with this project. 11. Waljee JF et al (2008) Effect of esthetic outcome after breast- conserving surgery on psychosocial functioning and quality of Funding Royal College Surgeons of England one year Research Fel- life. J Clin Oncol 26(20):3331–3337 lowship (awarded to ROC). The Royal Marsden/Institute of Cancer 12. Heil J et al (2010) Aesthetic and functional results after breast Research is an NIHR Biomedical Research Centre. This support is conserving surgery as correlates of quality of life measured by a acknowledged. German version of the Breast Cancer Treatment Outcome Scale (BCTOS). Breast 19(6):470–474 Compliance with ethical standards 13. Henseler H et al (2011) Investigation into accuracy and repro- ducibility of a 3D breast imaging system using multiple stereo cameras. J Plast Reconstr Aesthet Surg 64(5):577–582 Conflict of interest There are no conflicts of interest. 14. Kovacs L et al (2006) New aspects of breast volume measure- ment using 3-dimensional surface imaging. Ann Plast Surg Ethical approval All procedures performed in studies involving human 57(6):602–610 participants were in accordance with the ethical standards of the insti- 15. Cochrane RA et al (2003) Cosmesis and satisfaction after breast- tutional and/or national research committee and with the 1964 Helsinki conserving surgery correlates with the percentage of breast vol- Declaration and its later amendments or comparable ethical standard. ume excised. Br J Surg 90(12):1505–1509 The protocol was reviewed and passed by Yorkshire and the Humber— 16. Sigurdson LJ, Kirkland SA (2006) Breast volume determination South Yorkshire. REC reference 14YH0175. in breast hypertrophy: an accurate method using two anthropo- morphic measurements. Plast Reconstr Surg 118(2):313–320 Informed consent Informed consent was obtained from all individual 17. Mailey B et al (2013) Clinical accuracy and reproducibility of participants included in the study. Additional informed consent was Portrait 3D Surgical Simulation Platform in breast augmentation. obtained from the individual participant for whom identifying informa- Aesthet Surg J 33(1):84–92 tion is included in this article. 18. Losken A et al (2005) Validating three-dimensional imaging of the breast. Ann Plast Surg 54(5):471–476 Open Access This article is distributed under the terms of the Crea- 19. Onesti MG et al (2004) Breast asymmetry: a new vision of this tive Commons Attribution 4.0 International License (http://creat iveco malformation. Acta Chir Plast 46(1):8–11 mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- 20. Henseler H et  al (2013) Subjective versus objective assess- tion, and reproduction in any medium, provided you give appropriate ment of breast reconstruction. J Plast Reconstr Aesthet Surg credit to the original author(s) and the source, provide a link to the 66(5):634–639 Creative Commons license, and indicate if changes were made. 21. Henseler H et al (2012) Objective evaluation of the latissimus dorsi flap for breast reconstruction using three-dimensional imag- ing. J Plast Reconstr Aesthet Surg 65(9):1209–1215 22. Moyer HR et al (2008) Three-dimensional digital evaluation of References breast symmetry after breast conservation therapy. J Am Coll Surg 207(2):227–232 1. Burke PH et al (1983) Stereophotographic measurement of change 23. Liu C et al (2010) The role of three-dimensional scanning tech- in facial soft tissue morphology following surgery. Br J Oral Surg nique in evaluation of breast asymmetry in breast augmentation: 21(4):237–245 a 100-case study. Plast Reconstr Surg 126(6):2125–2132 2. Burke PH, Beard LF, Stereo-photogrammetry of the face. Rep 24. Losken A et al (2005) An objective evaluation of breast symmetry Congr Eur Orthod Soc, 1967: 279–293 and shape differences using 3-dimensional images. Ann Plast Surg 3. O’Connell RL et al (2015) Review of three-dimensional (3D) sur- 55(6):571–575 face imaging for oncoplastic, reconstructive and aesthetic breast 25. O’Connell RL et al (2017) The potential role of three-dimen- surgery. Breast 24(4):331–342 sional surface imaging as a tool to evaluate aesthetic outcome 4. Nahabedian MY, Galdino G (2003) Symmetrical breast recon- after Breast Conserving Therapy (BCT). Breast Cancer Res Treat struction: is there a role for three-dimensional digital photogra- 164(2):385–393 phy? Plast Reconstr Surg 112(6):1582–1590 5. O’Connell RL et al (2016) Initial experience of the BREAST- Q breast-conserving therapy module. Breast Cancer Res Treat 160(1):79–89 1 3

Journal

Breast Cancer Research and TreatmentSpringer 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