7 March 2024 in Frontiers in Artificial Intelligence
Objective: Automated surgical step recognition (SSR) using AI has been a catalyst in the “digitization” of surgery. However, progress has been limited to laparoscopy, with relatively few SSR tools in endoscopic surgery. This study aimed to create a SSR model for transurethral resection of bladder tumors (TURBT), leveraging a novel application of transfer learning to reduce video dataset requirements.Materials and methods: Retrospective surgical videos of TURBT were manually annotated with the following steps of surgery: primary endoscopic evaluation, resection of bladder tumor, and surface coagulation. Manually annotated videos were then utilized to train a novel AI computer vision algorithm to perform automated video annotation of TURBT surgical video, utilizing a transfer-learning technique to pre-train on laparoscopic procedures. Accuracy of AI SSR was determined by comparison to human annotations as the reference standard.Results: A total of 300 full-length TURBT videos (median 23.96 min; IQR 14.13–41.31 min) were manually annotated with sequential steps of surgery. One hundred and seventy-nine videos served as a training dataset for algorithm development, 44 for internal validation, and 77 as a separate test cohort for evaluating algorithm accuracy. Overall accuracy of AI video analysis was 89.6%. Model accuracy was highest for the primary endoscopic evaluation step (98.2%) and lowest for the surface coagulation step (82.7%).Conclusion: We developed a fully automated computer vision algorithm for high-accuracy annotation of TURBT surgical videos. This represents the first application of transfer-learning from laparoscopy-based computer vision models into surgical endoscopy, demonstrating the promise of this approach in adapting to new procedure types.
11 September 2024 in BJUI Compass
Background: Neoadjuvant platinum‐based chemotherapy offers a modest survival advantage in muscle‐invasive bladder cancer (MIBC) for patients with pathologic response. B7‐H3 (CD276), an immune checkpoint overexpressed in various cancers, including urothelial‐cell carcinoma (UCC), has been associated with chemoresistance and poor oncologic outcomes. We aimed to explore if B7H3 expression on bladder biopsy samples was a predictive biomarker for pathologic response to neoadjuvant platinum‐based chemotherapy.Methods: This was a retrospective cohort study among MIBC patients receiving neoadjuvant platinum‐based chemotherapy followed by radical cystectomy. All patients underwent routine preoperative biopsy of their tumour. Immunohistochemistry was used to evaluate B7‐H3 expression from pre‐operative specimens. The primary outcome of interest was pathologic complete response (pCR). Statistical analysis included Mann–Whitney U test, Fisher's exact test, Kaplan–Meier method, and Cox regression for survival analysis.Results: Among 87 patients analysed, high B7‐H3 expression was found in 44.8% (n = 39) of patients. The median follow‐up periods were similar between the high and low B7‐H3 groups (high expression; 4.29 years [SD 3.04], low expression 3.94 years [SD 3.04], p = 0.60). Only 20.5% of patients with high B7‐H3 expression achieved pCR, compared to 41.7% in the low expression group (p = 0.04). Cox regression showed no significant differences in recurrence‐free or cancer‐specific survival between the high and low B7‐H3 expression groups (p > 0.05).Conclusion: High B7‐H3 expression is associated with a reduced likelihood of achieving pCR in MIBC patients undergoing neoadjuvant chemotherapy. This suggests B7‐H3's potential as a predictive biomarker for chemotherapy response. Further research is needed to explore the role of B7‐H3 on platinum‐based chemotherapy response in urothelial cancer.