SutureAgent: Learning Surgical Trajectories via Goal-conditioned Offline RL in Pixel Space

1. 1Learn surgical trajectories from endoscopic video using goal-conditioned offline RL
2. 2Model sequential dependency among adjacent motion steps to improve prediction accuracy
3. 3Use sparse waypoint annotations to provide sufficient supervision for the model
4. 4Evaluate the performance of SutureAgent in simulated and real-world surgical scenarios

This research benefits AI engineers and roboticists working on surgical robotics, as it enables more accurate and safe motion execution in robot-assisted suturing. The findings can be applied by machine learning researchers and engineers to improve the performance of surgical robots.

💡 Goal-conditioned offline RL can effectively learn surgical trajectories from endoscopic video, improving the accuracy and safety of robot-assisted suturing

💡 SutureAgent: Learning surgical trajectories via goal-conditioned offline RL in pixel space 🤖