1. Department of Surgery and Cancer, Imperial College London

11 The current and future use of imaging in urological robotic surgery: A survey of the2 European Association of Robotic Urological Surgeons3 Archie Hughes-Hallett, MRCS1, Erik K Mayer, PhD1, Philip Pratt, PhD2, Alex Mottrie, PhD3,4, Ara4 Darzi, FRS1,2, Justin Vale, MS1, 5 1. Department of Surgery and Cancer, Imperial College London 6 2. The Hamlyn Centre for Robotic Surgery, Imperial College London 7 3. Department of Urology, OLV Clinic, Aalst, Belgium 8 4. O.L.V. Vattikuti Robotic Surgery Institute, Aalst, Belgium 9 1011 Corresponding Author 12 Erik Mayer, 13 Department of Surgery and Cancer, Imperial College London, St Marys Hospital14 Campus, London, W2 1NY15 0798419564216 [email protected]1718 No reprints will be available from the authors1920 No financial support was received2122 Article Category: Original Article2324 Word count abstract: 24425 Word count manuscript text: 2,14226 5 figures and 2 tables2728 22930 Introduction31 Since Röntgen first utilised x-rays to image the carpal bones of the human hand in32 1895, medical imaging has evolved and is now able to provide a detailed33 representation of a patient’s intracorporeal anatomy, with recent advances now34 allowing for 3-dimensional (3D) reconstructions. The visualisation of anatomy in 3D35 has been shown to improve the ability to localize structures when compared to 2D36 with no change in the amount of cognitive loading [1]. This has allowed imaging to37 move from a largely diagnostic tool to one that can be used for both diagnosis and38 operative planning. 3940 One potential interface to display 3D images, to maximise its potential as a tool for41 surgical guidance, is to overlay them onto the endoscopic operative scene (augmented42 reality). This addresses, in part, a criticism often levelled at robotic surgery, the loss43 of haptic feedback. Augmented reality has the potential to mitigate for this sensory44 loss by enhancing the surgeons visual cues with information regarding subsurface45 anatomical relationships [2].4647 Augmented reality surgery is in its infancy for intra-abdominal procedures due in48 large part to the difficulties of applying static preoperative imaging to a constantly49 deforming intraoperative scene [3]. There are case reports and ex-vivo studies in the50 literature examining the technology in minimal access prostatectomy [3–6] and51 partial nephrectomy [7–10], but there remains a lack of evidence determining52 whether surgeons feel there is a role for the technology and if so what procedures53 they feel it would be efficacious for.2 35455 This questionnaire-based study was designed to assess: firstly, the pre- and56 intraoperative imaging modalities utilised by robotic urologists; secondly, the current57 use of imaging intraoperatively for surgical planning; and finally whether there is a58 desire for augmented reality amongst the robotic urological community. 5960 Methods61 Recruitment62 A web based survey instrument was designed and sent out, as part of a larger survey,63 to members of the EAU Robotic Urology Section (ERUS). Only independently64 practising robotic surgeons performing RALP, RAPN and/or robotic cystectomy65 were included in the analysis, those surgeons exclusively performing other66 procedures were excluded. Respondents were offered no incentives to reply. All data67 collected was anonymous. 6869 Survey design and administration70 The questionnaire was created using the LimeSurvey platform71 (www.limesurvey.com) and hosted on their website. All responses (both complete72 and incomplete) were included in the analysis. The questionnaire was dynamic with73 the questions displayed tailored to the respondents’ previous answers.7475 When computing fractions or percentages the denominator was the number of76 respondents to answer the question, this number is variable due to the dynamic nature77 of the questionnaire.78 479 Survey Content80 Demographics81 All respondents to the survey were asked in what country they practised and what82 robotic urological procedures they performed, in addition to what procedures they83 performed surgeons were asked specify the number of cases they had undertaken for84 each procedure. 8586 Current Imaging Practice87 Procedure-specific questions in this group were displayed according to the operations88 the respondent performed. A summary of the questions can be seen in appendix 1.89 Procedure non-specific questions were also asked. Participants were asked whether90 they routinely used the Tile Pro™ function of the da Vinci console (Intuitive91 Surgical, Sunnyvale, USA) and whether they routinely viewed imaging92 intraoperatively.9394 Augmented Reality95 Prior to answering questions in this section, participants were invited to watch a96 video demonstrating an augmented reality platform during Robot-Assisted Partial97 Nephrectomy (RAPN), performed by our group at Imperial College London. A still98 from this video can be seen in figure 1. They were then asked whether they felt99 augmented reality would be of use as a navigation or training tool in robotic surgery.100101 Once again, in this section, procedure-specific questions were displayed according to102 the operations the respondent performed. Only those respondents who felt augmented103 reality would be of use as a navigation tool were asked procedure-specific questions.4 5104 Questions were asked to establish where in these procedures they felt an augmented105 reality environment would be of use.106107 Results108 Demographics109 Of the 239 respondents completing the survey 117 were independently practising110 robotic surgeons and were therefore eligible for analysis. The majority of the111 surgeons had both trained (210/239, 87.9%) and worked in Europe (215/239, 90.0%).112 The median number of cases undertaken by those surgeons reporting their case113 volume was: 120 (6 - 2000), 9 (1 – 120) and 30 (1 – 270), for RALP, Robot assisted114 cystectomy and RAPN respectively.115116 Contemporary use of imaging in robotic surgery117 When enquiring about the use of imaging for surgical planning, the majority of118 surgeons (57%, 65/115) routinely viewed preoperative imaging intraoperatively with119 only 9% (13/137) routinely capitalising on the TilePro™ function in the console to120 display these images, when assessing the use of TilePro™ amongst surgeons who121 performed RAPN 13.8% (9/65) reported using the technology routinely.122123 When assessing the imaging modalities that are available to a surgeon in theatre the124 majority of surgeons performing RALP (74%, 78/106)) reported using MRI with an125 additional 37% (39/106) reporting the use of CT for preoperative staging and/or126 planning. For surgeons performing RAPN and robot-assisted cystectomy there was127 more of a consensus with 97% (68/70) and 95% (54/57) of surgeons, respectively,128 using CT for routine preoperative imaging (table 1). 6129130 Those surgeons performing RAPN were found to have the most diversity in the way131 they viewed preoperative images in theatre, routinely viewing images in sagittal,132 coronal and axial slices (table 2). The majority of these surgeons also viewed the133 images as 3D reconstructions (54%, 38/70). 134135 The majority of surgeons used ultrasound intraoperatively in RAPN (51%, 35/69)136 with a further 25% (17/69) reporting they would use it if they had access to a ‘drop-137 in’ ultrasound probe (figure 3).138139 Desire for augmented reality140 In all 87% of respondents envisaged a role for augmented reality as a navigation tool141 in robotic surgery and 82% (88/107) felt that there was an additional role for the142 technology as a training tool.143144 The greatest desire for augmented reality was amongst those surgeons performing145 RAPN with 86% (54/63) feeling the technology would be of use. The largest group146 of surgeons felt it would be useful in identifying tumour location, with significant147 numbers also feeling it would be efficacious in tumour resection (figure 4).148149 When enquiring about the potential for augmented reality in Robot-Assisted150 Laparoscopic Prostatectomy (RALP), 79% (20/96) of respondents felt it would be of151 use during the procedure, with the largest group feeling it would be helpful for nerve152 sparing 65% (62/96) (Figure 2). The picture in cystectomy was similar with 74%153 (37/50) of surgeons believing augmented reality would be of use, with both nerve6 7154 sparing and apical dissection highlighted as specific examples (40%, 20/50) (Figure155 5). The majority also felt that it would be useful for lymph node dissection in both156 RALP and robot assisted cystectomy (55% (52/95) and 64% (32/50) respectively). 157158 Discussion159 The results from this study suggest that the contemporary robotic surgeon views160 imaging as an important adjunct to operative practice. The way these images are161 being viewed is changing; although the majority of surgeons continue to view images162 as two-dimensional (2D) slices a significant minority have started to capitalise on 3D163 reconstructions to give them an improved appreciation of the patient’s anatomy. 164165 This study has highlighted surgeons’ willingness to take the next step in the166 utilisation of imaging in operative planning, augmented reality, with 87% feeling it167 has a role to play in robotic surgery. Although there appears to be a considerable168 desire for augmented reality, the technology itself is still in its infancy with the169 limited evidence demonstrating clinical application reporting only qualitative results170 [3,11–13].171172 There are a number of significant issues that need to be overcome before augmented173 reality can be adopted in routine clinical practice. The first of these is registration (the174 process by which two images are positioned in the same coordinate system such that175 the locations of corresponding points align [14]). This process has been performed176 both manually and using automated algorithms with varying degrees of accuracy177 [2,15]. The second issue pertains to the use of static preoperative imaging in a 8178 dynamic operative environment; in order for the preoperative imaging to be179 accurately registered it must be deformable. This problem remains as yet unresolved.180181 Live intraoperative imaging circumvents the problems of tissue deformation and in182 RAPN 51% of surgeons reported already using intraoperative ultrasound to aid in183 tumour resection. Cheung and colleagues [9] have published an ex-vivo study184 highlighting the potential for intraoperative ultrasound in augmented reality partial185 nephrectomy. They report the overlaying of ultrasound onto the operative scene to186 improve the surgeon’s appreciation of the subsurface tumour anatomy, this187 improvement in anatomical appreciation resulted in improved resection quality over188 conventional ultrasound guided resection [9]. Building on this work the first in vivo189 use of overlaid ultrasound in RAPN has recently been reported [10]. Although good190 subjective feedback was received from the operating surgeon, the study was limited191 to a single case demonstrating feasibility and as such was not able to show an192 outcome benefit to the technology [10].193194 RAPN also appears to be the area in which augmented reality would be most readily195 adopted with 86% of surgeons claiming they see a use for the technology during the196 procedure. Within this operation there are two obvious steps to augmentation,197 anatomical identification (in particular vessel identification to facilitate both routine198 ‘full clamping’ and for the identification of secondary and tertiary vessels for199 ‘selective clamping’ [16]) and tumour resection. These two phases have different200 requirements from an augmented reality platform; the first phase of identification201 requires a gross overview of the anatomy without the need for high levels of202 registration accuracy. Tumour resection, however, necessitates almost sub-millimetre8 9203 accuracy in registration and needs the system to account for the dynamic204 intraoperative environment. The step of anatomical identification is amenable to the205 use of non-deformable 3D reconstructions of preoperative imaging while that of206 image-guided tumour resection is perhaps better suited to augmentation with live207 imaging such as ultrasound [2,9,17].208209 For RALP and robot-assisted cystectomy the steps in which surgeons felt augmented210 reality would be of assistance were those of neurovascular bundle preservation and211 apical dissection. The relative, perceived, efficacy of augmented reality in these steps212 correlate with previous examinations of augmented reality in RALP [18,19].213 Although surgeon preference for utilising AR while undertaking robotic214 prostatectomy has been demonstrated, Thompson et al. failed to demonstrate an215 improvement in oncological outcomes in those patients undergoing AR RALP [19]. 216217 Both nerve sparing and apical dissection require a high level of registration accuracy218 and a necessity for either live imaging or the deformation of preoperative imaging to219 match the operative scene; achieving this level of registration accuracy is made more220 difficult by the mobilisation of the prostate gland during the operation [18]. These221 problems are equally applicable to robot-assisted cystectomy. Although guidance222 systems have been proposed in the literature for RALP [3,4,13,18,20], none have223 achieved the level of accuracy required to provide assistance during nerve sparing.224 Additionally, there are still imaging challenges that need to be overcome. Although225 multiparametric MRI has been shown to improve decision making in opting for a226 nerve sparing approach to RALP [21] the imaging is not yet able to reliably discern227 the exact location of the neurovascular bundle. This said significant advances are 10228 being made with novel imaging modalities on the horizon that may allow for imaging229 of the neurovascular bundle in the near future [22].230231 Limitations232233 The number of operations included represents a significant limitation of the study,234 had different index procedures been chosen different results may have been seen.235 This being said the index procedures selected were chosen as they represent the vast236 majority of uro-oncological robotic surgical practice, largely mitigating for this237 shortfall.238239 Although the available ex-vivo evidence suggests that introducing augmented reality240 operating environments into surgical practice would help to improve outcomes [9,23]241 the in-vivo experience to date is limited to small volume case series reporting242 feasibility [2,3,15]. To date no study has demonstrated an in-vivo outcome advantage243 to augmented reality guidance. In addition to this limitation augmented reality has244 been demonstrated to increased rates of inattention blindness amongst surgeons245 suggesting there is a trade of between increasing visual information and the surgeon’s246 ability to appreciate unexpected operative events [23].247248 Conclusions249250 This survey depicts the contemporary robotic surgeon to be comfortable with the use251 of imaging to aid in intraoperative planning; furthermore it highlights a significant252 interest amongst the urological community in augmented reality operating platforms.10 11253254 Short to medium term development of augmented reality systems in robotic urology255 surgery would be best performed using RAPN as the index procedure. Not only was256 this the operation where surgeons saw the greatest potential benefits, but it may also257 be the operation where it is most easily achievable by capitalising on the respective258 benefits of technologies the surgeons are already using; preoperative CT for259 anatomical identification and intraoperative ultrasound for tumour resection.260261 Conflicts of Interest262 None of the authors have any conflicts of interest to declare263264 References265 1. Foo J-L, Martinez-Escobar M, Juhnke B, Cassidy K, Hisley K, Lobe T, Winer266 E. Evaluating mental workload of two-dimensional and three-dimensional267 visualization for anatomical structure localization. J Laparoendosc Adv Surg268 Tech A. 2013;23(1):65–70. 269 2. Hughes-Hallett A, Mayer EK, Marcus HJ, Cundy TP, Pratt PJ, Darzi AW,270 Vale J. Augmented Reality Partial Nephrectomy: Examining the Current271 Status and Future Perspectives. Urology; 2014;83(2):266–73. 272 3. Sridhar AN, Hughes-Hallett A, Mayer EK, Pratt PJ, Edwards PJ, Yang G-Z,273 Darzi AW, Vale J. Image-guided robotic interventions for prostate cancer. Nat274 Rev Urol. 2013;10(8):452–62. 275 4. Cohen D, Mayer E, Chen D, Anstee A, Vale J, Yang G-Z, Darzi A, Edwards P276 “Eddie.” Augmented reality image guidance in minimally invasive277 prostatectomy. Lect Notes Comput Sci. 2010;6367:101–10. 12278 5. Simpfendorfer T, Baumhauer M, Muller M, Gutt CN, Meinzer HP, Rassweiler279 JJ, Guven S, Teber D. Augmented reality visualization during laparoscopic280 radical prostatectomy. J Endourol. 2011;25(12):1841–5. 281 6. Teber D, Simpfendorfer T, Guven S, Baumhauer M, Gozen AS, Rassweiler J.282 In-vitro evaluation of a soft-tissue navigation system for laparoscopic283 prostatectomy. J Endourol. 2010;24(9):1487–91. 284 7. Teber D, Guven S, Simpfendörfer T, Baumhauer M, Güven EO, Yencilek F,285 Gözen AS, Rassweiler JJ, Simpfendorfer T, Guven EO, Gozen AS.286 Augmented reality: a new tool to improve surgical accuracy during287 laparoscopic partial nephrectomy? Preliminary in vitro and in vivo results. Eur288 Urol. 2009;56(2):332–8. 289 8. Pratt P, Mayer E, Vale J, Cohen D, Edwards E, Darzi A, Yang G-Z. An290 effective visualisation and registration system for image-guided robotic partial291 nephrectomy. J Robot Surg. 2012;6(1):23–31. 292 9. Cheung CL, Wedlake C, Moore J, Pautler SE, Peters TM. Fused video and293 ultrasound images for minimally invasive partial nephrectomy: A phantom294 study. Med Image Comput Comput Assist Interv. 2010;13(Pt 3):408–15. 295 10. Hughes-Hallett A, Pratt P, Mayer E, Di Marco A, Yang G-Z, Vale J, Darzi A.296 Intraoperative Ultrasound Overlay in Robot-assisted Partial Nephrectomy:297 First Clinical Experience. Eur Urol. 2013; 298 11. Nakamura K, Naya Y, Zenbutsu S, Araki K, Cho S, Ohta S, Nihei N, Suzuki299 H, Ichikawa T, Igarashi T. Surgical navigation using three-dimensional300 computed tomography images fused intraoperatively with live video. J301 Endourol. 2010;24(4):521–4. 12 13302 12. Teber D, Guven S, Simpfendorfer T, Baumhauer M, Guven EO, Yencilek F,303 Gozen AS, Rassweiler J. Augmented Reality: A New Tool To Improve304 Surgical Accuracy during Laparoscopic Partial Nephrectomy? Preliminary In305 Vitro and In Vivo Results. Eur Urol. 2009;56(2):332–8. 306 13. Ukimura O, Gill IS. Imaging-assisted endoscopic surgery: Cleveland clinic307 experience. J Endourol. 2008;22(4):803–9. 308 14. Altamar HO, Ong RE, Glisson CL, Viprakasit DP, Miga MI, Herrell SD,309 Galloway RL. Kidney deformation and intraprocedural registration: A study of310 elements of image-guided kidney surgery. J Endourol. 2011;25(3):511–7. 311 15. Nicolau S, Soler L, Mutter D, Marescaux J. Augmented reality in laparoscopic312 surgical oncology. Surg Oncol. 2011;20(3):189–201. 313 16. Ukimura O, Nakamoto M, Gill IS. Three-dimensional reconstruction of314 renovascular-tumor anatomy to facilitate zero-ischemia partial nephrectomy.315 Eur Urol. 2012;61(1):211–7. 316 17. Pratt P, Hughes-Hallett A, Di Marco A, Cundy T, Mayer E, Vale J, Darzi A,317 Yang G-Z. Multimodal Reconstruction for Image-Guided Interventions.318 Hamlyn Symposium. 2013. 319 18. Mayer EK, Cohen D, Chen D, Anstee A, Vale J a., Yang GZ, Darzi AW,320 Edwards E. Augmented Reality Image Guidance in Minimally Invasive321 Prostatectomy. Eur Urol Supp. 2011;10(2):300. 322 19. Thompson S, Penney G, Billia M, Challacombe B, Hawkes D, Dasgupta P.323 Design and evaluation of an image-guidance system for robot-assisted radical324 prostatectomy. BJU Int. 2013;111(7):1081–90. 14325 20. Simpfendorfer T, Baumhauer M, Muller M, Gutt CN, Meinzer H-PP,326 Rassweiler JJ, Guven S, Teber D, Simpfendörfer T, Müller M. Augmented327 reality visualization during laparoscopic radical prostatectomy. J Endourol.328 2011;25(12):1841–5. 329 21. Panebianco V, Salciccia S, Cattarino S, Minisola F, Gentilucci A, Alfarone A,330 Ricciuti GP, Marcantonio A, Lisi D, Gentile V, Passariello R, Sciarra A. Use331 of Multiparametric MR with Neurovascular Bundle Evaluation to Optimize the332 Oncological and Functional Management of Patients Considered for Nerve-333 Sparing Radical Prostatectomy. J Sex Med. 2012;9(8):2157–66. 334 22. Rai S, Srivastava A, Sooriakumaran P, Tewari A. Advances in imaging the335 neurovascular bundle. Curr Opin Urol. 2012;22(2):88–96. 336 23. Dixon BJ, Daly MJ, Chan H, Vescan AD, Witterick IJ, Irish JC. Surgeons337 blinded by enhanced navigation: the effect of augmented reality on attention.338 Surg Endosc. 2013;27(2):454–61. 33934034134234334434534634734814 15349350351352353354355 TablesCT MRI USS None Other RALP (n=106) 39.8% 73.5% 2% 15.1% 8.4% (39) (78) (3) (16) (9) RAPN (n=70) 97.1% 42.9% 17.1% 0% 2.9% (68) (30) (12) (0) (2) Cystectomy (n=57) 94.7% 26.3% 1.8% 1.8% 5.3% (54) (15) (1) (1) (3) Table 1 - Which preoperative imaging modalities do you use for diagnosis and surgical planning? 356357358359360361362363364365366 16367368Axial Coronal Sagittal 3D Do not slices slices slices (n) recons. view (n) (n) (n) (n) RALP (n=106) 49.1% 44.3% 31.1% 9.4% 31.1% (52) (47) (33) (10) (33) RAPN (n=70) 68.6% 74.3% 60% (42) 54.3% 0% (48) (52) (38) (0) Cystectomy 70.2% 52.6% 50.9% 21.1% 8.8% (n=57) (40) (30) (29) (12) (5) Table 2 - How do you typically view preoperative imaging in the OR? 3D recons = Three dimensional reconstructions 36937037137237337437537637737837938016 17381 Figure Legends382383 Figure 1 – A still taken from a video of augmented reality robot assisted partial384 nephrectomy performed. Here the tumour has been painted into the operative view385 allowing the surgeon to appreciate the relationship of the tumour to the surface of the386 kidney.387 Figure 2 – Chart demonstrating responses to the question - In robotic prostatectomy 388 which parts of the operation do you feel augmented reality image overlay would be of389 assistance?390391 Figure 3 - Chart demonstrating responses to the question - Do you use intraoperative392 ultrasound for robotic partial nephrectomy?393394 Figure 4 - Chart demonstrating responses to the question – In robotic partial395 nephrectomy which parts of the operation do you feel augmented reality image396 overlay would be of assistance?397398 Figure 5 - Chart demonstrating responses to the question – In robotic cystectomy399 which parts of the operation do you feel augmented reality overlay technology would400 be of assistance?

1. Department of Surgery and Cancer, Imperial College London

Details

Download

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

Support