Vision-Based Force Estimation for Robotic Surgery
Estimating contact forces in robotic surgery without a force sensor (Max-Planck Institute).
As a visiting researcher at the Max-Planck Institute for Intelligent Systems (advisor: Katherine J. Kuchenbecker), I built models that recover contact forces in robotic surgery from indirect signals.
- Deep neural networks estimate palpation forces during robotic surgery from visual and inertial (IMU) data, without any direct force or haptic sensor.
- Recurrent and convolutional models trained on noisy, multimodal visual-inertial streams from a da Vinci surgical instrument to regress contact forces (IEEE Transactions on Medical Robotics and Bionics, 2023; SAGES, 2020).