Robots improve control of deep soft tissue with causal sensing
CADeT: Causal-Aware Deformation Transmission for Indirect Robotic Manipulation of Soft Tissue
Robotics
Summary
Indirectly moving soft tissues deep inside the body during robot-assisted surgery is hard because tissues between the robot and target tissue change how forces and movements travel. The authors created a new way called CADeT that helps the robot understand these hidden changes by using cause-effect models and special probing moves. This approach lets the robot guess what kind of tissue connection mode is active and learn how forces transmit in real time, improving the accuracy of controlling tissue shapes. Tests in both simulations and real tissue experiments show this method works better and faster than older methods.
What this means in practice
- •For robotic surgery teams: Enhance control precision in robot-assisted surgeries by interpreting tissue deformation transmission modes and adjusting manipulation accordingly.
- •For medical robotics developers: Incorporate causal-aware sensing algorithms to improve real-time feedback and control of robotic tools manipulating inaccessible soft tissues.
Authors
Junlei Hu, Dominic Jones, Pietro Valdastri
Abstract
Indirect manipulation of deep-seated deformable anatomy inaccessible to the robot is challenging in robot-assisted minimally invasive surgery (RAMIS) because intervening tissues spatially filter deformation transmission. Passive observations can be ambiguous because the Decoupled and Blocked modes may produce similar motion responses. We propose CADeT, a causal-aware deformation transmission framework that integrates structural causal model (SCM) with active sensing to infer a latent transmission mode and estimate a state-dependent adhesion Jacobian online. During normal manipulation, control actions update the mode belief; when ambiguity persists, an additional probing action is selected to improve mode distinguishability. The mode belief and learned Jacobian are incorporated into a belief-aware model predictive controller for indirect target-shape control. Validation in simulation and on the da Vinci research kit (dVRK), using phantom and ex vivo porcine tissues, shows higher mode-identification accuracy and faster shape-error convergence than the evaluated model-free and model-based baselines. These results show that active sensing improves mode identification and indirect deformation control under the evaluated conditions.