Imitation learning for supervised autonomous tumor resection in central airway obstruction

Central airway obstruction (CAO) procedures require precise, repetitive dual-arm coordination under progressively degraded visibility from smoke and charring. Model-based control struggles with deformable-tissue variability and real-time constraints, while existing data-driven methods emphasize short, single-skill tasks rather than cyclic workflows. We investigate whether imitation learning can achieve fully autonomous task-level and high-level for CAO tumor resection on da Vinci Research Kit.

We propose a hierarchical framework combining low-level action chunking with transformers (ACT) policies for autonomous task-level execution with finite state machine (FSM)-based high-level coordination for procedure-level sequencing. The workflow decomposes into five recurring tasks. Policies are trained on synchronized multi-view video and kinematics with hybrid-relative actions. We also analyze task-level policies with different data regimes through an ablation study.

Low-level policies achieved fully autonomous execution with 0.938mm RMSE without intervention, comparable to an operator (0.968 mm) and surgeon (1.086 mm). High-level coordination enabled supervised autonomous full procedure execution with minimal interventions resulting in 1.232mm accuracy. Ablation studies showed balanced task coverage with minimal data (3%) maintained coherent behavior (1.614mm), while random 60% sampling yielded 1.386mm. Blinded evaluation by expert surgeons confirmed comparable resection quality and superior retraction performance relative to human operators.

This demonstrates the first learning-based supervised autonomy for CAO removal procedure with expert-level accuracy. The hierarchical framework enables full procedure execution requiring supervision only at task transitions. We publicly release the training datasets including multi-view imagery and robot kinematics as a benchmark for surgical automation research.