Artificial-Intelligence-based Surgical Phase Recognition in Robot-Assisted Radical Prostatectomy and Cross-Surgeon Validation

Comprehensive Summary

With continued advancements in artificial intelligence (AI), the practical application of AI in surgery still remains relatively limited. Using clinical data from surgical operations performed by six different experienced surgeons, an AI system capable of automatically recognizing surgical phases in robot-assisted radical prostatectomy (RARP) was developed. The basis of this AI system was the classification of surgical operations into nine different phases and the incorporation of Temporal Convolutional Networks for the Operating Room (TeCNO) for model development and precision analysis. Testing of this model revealed not only high accuracy, but also demonstrated generalizability across different surgeons, providing valuable feedback for the continued development of AI incorporation into surgical procedures.

Outcomes and Implications

The development of AI that has the ability to recognize various phases of a surgery as well as assess surgery quality and provide feedback gives the opportunity to greatly improve patient outcomes and surgical education. By gradually incorporating such a tool into medical practice, physicians and surgeons will be able to quickly identify areas within their practice that can be improved and adjusted to maximize efficiency and effectiveness of procedures.