AI-powered real-time annotations during urologic surgery: The future of training and quality metrics.

INTRODUCTION AND OBJECTIVE: Real-time artificial intelligence (AI) annotation of the surgical field has the potential to automatically extract information from surgical videos, helping to create a robust surgical atlas. This content can be used for surgical education and qualitative initiatives. We demonstrate the first use of AI in urologic robotic surgery to capture live surgical video and annotate key surgical steps and safety milestones in real-time. SUMMARY BACKGROUND DATA: While AI models possess the capability to generate automated annotations based on a collection of video images, the real-time implementation of such technology in urological robotic surgery to aid surgeon and training staff it is still pending to be studied. METHODS: We conducted an educational symposium, which broadcasted 2 live procedures, a robotic-assisted radical prostatectomy (RARP) and a robotic-assisted partial nephrectomy (RAPN). A surgical AI platform system (Theator, Palo Alto, CA) generated real-time annotations and identified operative safety milestones. This was achieved through trained algorithms, conventional video recognition, and novel Video Transfer Network technology which captures clips in full context, enabling automatic recognition and surgical mapping in real-time. RESULTS: Real-time AI annotations for procedure #1, RARP, are found in Table 1. The safety milestone annotations included the apical safety maneuver and deliberate views of structures such as the external iliac vessels and the obturator nerve. Real-time AI annotations for procedure #2, RAPN, are found in Table 1. Safety milestones included deliberate views of structures such as the gonadal vessels and the ureter. AI annotated surgical events included intraoperative ultrasound, temporary clip application and removal, hemostatic powder application, and notable hemorrhage. CONCLUSIONS: For the first time, surgical intelligence successfully showcased real-time AI annotations of 2 separate urologic robotic procedures during a live telecast. These annotations may provide the technological framework for send automatic notifications to clinical or operational stakeholders. This technology is a first step in real-time intraoperative decision support, leveraging big data to improve the quality of surgical care, potentially improve surgical outcomes, and support training and education.

Transperitoneal Versus Retroperitoneal Robotic-Assisted Partial Nephrectomy in Patients with Obesity.

Introduction: We aim to compare transperitoneal (TP) and retroperitoneal (RP) robotic partial nephrectomy (RPN) in obese patients. Obesity and RP fat can complicate RPN, especially in the RP approach where working space is limited. Materials and Methods: Using a multi-institutional database, we analyzed 468 obese patients undergoing RPN for a renal mass (86 [18.38%] RP, 382 [81.62%] TP). Obesity was defined as body mass index ≥30 kg/m2*. A 1:1 propensity score matching was performed adjusting for age, previous abdominal surgery, tumor size, R.E.N.A.L nephrometry score, tumor location, surgical date, and participating centers. Baseline characteristics and perioperative and postoperative data were compared. Results: In the propensity score-matched cohort, 79 (50%) TP patients were matched with 79 (50%) RP patients. The RP group had more posterior tumors (67 [84.81%], RP versus 23 [29.11%], TP; P < .001), while the other baseline characteristics were comparable. Warm ischemia time (interquartile range; 15 [10, 12], RP versus 14 [10, 17] minutes, TP; P = .216), operative time (129 [116, 165], RP versus 130 [95, 180] minutes, TP; P = .687), estimated blood loss (50 [50, 100], RP versus 75 [50, 150] mL, TP; P = .129), length of stay (1 [1, 1], RP versus 1 [1, 2] day, TP; P = .319), and major complication rate (1 [1.27%], RP versus 3 [3.80%], TP; P = .620) were similar. No significant difference was observed in positive surgical margin rate and delta estimated glomerular filtration at follow-up. Conclusion: TP and RP RPN yielded similar perioperative and postoperative outcomes in obese patients. Obesity should not be a factor in determining optimal approach for RPN.

Robot-assisted partial nephrectomy for complex renal tumors: Analysis of a large multi-institutional database.

INTRODUCTION: Highly complex renal masses pose a challenge to urologic surgeons' ability to perform robotic partial nephrectomy (RPN). Given the increased utilization of the robotic approach for small renal masses, we sought to characterize the outcomes and determine the safety and feasibility of RPN for complex renal masses from our large multi-institutional cohort. METHODS: We performed a retrospective analysis of patients with R.E.N.A.L. Nephrometry Scores ≥10 who underwent RPN in our multi-institutional cohort (N = 372). Baseline demographic, clinical and tumor related characteristics were evaluated with the primary endpoint of trifecta achievement (defined as negative surgical margin, no major complications, and warm ischemia time ≤25 min). Relationships between variables were assessed using the chi-square test of independence, Fisher exact test, Mann-Whitney U test, and Kruskal Wallis test. Logistic regression was used to evaluate the relationship between baseline characteristics and trifecta achievement. RESULTS: Of 372 patients in the study, mean age was 58 years, and median BMI was 30.49 kg/m2. The median tumor size was 4.3 cm (3.0-5.9 cm). Most of the patients had R.E.N.A.L. scores of 10 (n = 253; 67.01%). Overall, trifecta was achieved in 72.04% of patients. Stratifying intraoperative and postoperative outcomes by R.E.N.A.L. scores, there was no significant difference in trifecta achievement, operative time, warm ischemia time (WIT), open conversion, major complication, or positive margin rates. Length of hospital stay was significantly longer for higher R.E.N.A.L. scores (median days 2 vs. 1, P = 0.012). Multivariate analyses for factors associated with trifecta achievement concluded that age and baseline eGFR were independently associated with trifecta achievement. CONCLUSION: RPN is a safe and reproducible procedure for complex tumors with R.E.N.A.L. Nephrometry scores ≥10. Our results suggest excellent rates of trifecta achievement and short-term functional outcomes when performed by experienced surgeons. Long-term oncological and functional evaluation are needed to further support this conclusion.

A comparison of outcomes between transperitoneal and retroperitoneal robotic assisted partial nephrectomy in patients with completely endophytic kidney tumors.

INTRODUCTION: Retroperitoneal robotic partial nephrectomy (RPN) has been shown to have comparable outcomes to the transperitoneal approach for renal tumors. However, this may not be true for completely endophytic tumors as they pose significant challenges in RPN with increased complication rates. Hence, we sought to compare the safety and feasibility of retroperitoneal RPN to transperitoneal RPN for completely endophytic tumors. METHODS: We performed a retrospective analysis of patients who underwent RPN for a completely endophytic renal mass using either transperitoneal or retroperitoneal approach from our multi-institutional database (n = 177). Patients who had a solitary kidney, prior ipsilateral surgery, multiple/bilateral tumors, and horseshoe kidneys were excluded from the analysis. Overall, 156 patients were evaluated (112 [71.8%] transperitoneal, 44 [28.2%] retroperitoneal). Baseline characteristics, perioperative and postoperative data were compared between the surgical transperitoneal and retroperitoneal approach using Chi-square test, Fishers exact test, t test, Mood median test and Mann Whitney U test. RESULTS: Of the 156 patients in this study, 86 (56.9%) were male and the mean (SD) age was 58 (13) years. Baseline characteristics were comparable between the 2 approaches. Compared to transperitoneal approach, retroperitoneal approach had similar ischemia time (19.6 [SD = 7.6] minutes vs. 19.5 [SD = 10.2] minutes, P = 0.952), operative time (157.5 [SD = 44.8] minutes vs. 160.2 [SD = 47.3] minutes, P = 0.746), median estimated blood loss (50 ml [IQR: 50, 150] vs. 100 ml [IQR: 50, 200], P = 0.313), median length of stay (1 [IQR: 1, 2] day vs. 1 [IQR: 1, 2] day, P = 0.126) and major complication rate (2 [4.6%] vs. 3 [2.7%], P = 0.621). No difference was observed in positive surgical margin rate (P = 0.1.00), delta eGFR (P = 0.797) and de novo chronic kidney disease occurrence (P = 1.000). CONCLUSION: Retroperitoneal and transperitoneal RPN yielded similar perioperative and functional outcomes in patients with completely endophytic tumors. In well-selected patients with purely endophytic tumors, either a retroperitoneal or transperitoneal approach could be considered without compromising perioperative and postoperative outcomes.