Surgical Gesture

Surgical gesture recognition and prediction are active research areas aiming to improve safety and efficiency in robot-assisted surgery. Current research focuses on developing robust and real-time models, often employing multimodal transformer architectures and temporal convolutional networks (TCNs), to analyze kinematic and video data for accurate gesture identification and prediction. These advancements enable improved surgical skill assessment, error detection, and ultimately, the development of more autonomous and safer surgical systems. The ability to generalize these models across different surgical tasks and individuals is a key challenge and focus of ongoing work.