Development of a machine learning-based tension measurement method in robotic surgery
Surgical Endoscopy, 2025
Anastomotic leaks complicate up to 10% of the 300,000 colorectal surgeries performed annually in the U.S., with tension being a key but subjectively assessed risk factor. This study evaluates the feasibility of an objective method to measure mechanical tension in ex vivo porcine colons using a machine learning algorithm integrated with the da Vinci Research Kit (dVRK). The algorithm estimated forces applied by robotic arms with high accuracy (up to 88%) and strong correlation (Spearman’s Correlation > 0.80) with ground-truth sensor data. This novel approach represents the first robotic measurement of tissue tension and has the potential to improve surgical outcomes by reducing anastomotic leaks.
Recommended citation: Khan A*, Yang H*, Habib DRS, Ali D, Wu JY. Development of a machine learning-based tension measurement method in robotic surgery. Surg Endosc. 2025. doi:10.1007/s00464-025-11658-9