Keshara Weerasinghe

Papers

Total Citations

H-Index

About

Keshara Weerasinghe is a rising researcher at the forefront of surgical robotics and artificial intelligence, specializing in real-time activity recognition and prediction for robot-assisted surgery. Their most-cited work introduces a multimodal transformer architecture that simultaneously predicts surgical gestures and instrument trajectories from short video segments, directly addressing critical challenges in intraoperative safety and autonomy. By fusing visual and kinematic data, Weerasinghe’s approach enables high-accuracy, low-latency predictions that could transform how surgical robots assist human surgeons, reducing errors and improving patient outcomes. With 9 citations already for their 2024 paper—a strong early indicator of impact—this work has quickly become a reference point in the growing field of surgical AI. Weerasinghe’s contributions bridge deep learning and clinical application, offering a practical path toward semi-autonomous robotic systems that learn from expert demonstrations. Their research not only advances technical frontiers but also holds promise for training next-generation surgeons and enhancing procedural consistency. As a young investigator, Weerasinghe is poised to shape the future of intelligent surgical environments.