Robotic Optical Coherence Tomography-Guided Geometric Modeling of Soft Tissue

3D Representations for Data-Driven Closed-Loop Shape Control

Chaoyu Zhang* Guangshen Ma*,† Mark Draelos

University of Michigan, Ann Arbor

* Equal contribution    † Project lead

Code

Abstract

In medical robotics, data-driven soft-tissue shape control requires deforming tissue toward desired morphologies through sequential physical interactions. This challenge is particularly important in microscale manipulation, where localized deformation is difficult to observe, represent, and control, while physics-based models can be computationally expensive and difficult to adapt to realistic environments. Optical coherence tomography (OCT) provides high-resolution volumetric perception of tissue deformation, but target-conditioned control requires effective representations of both observed and desired tissue states. In this work, we introduce an OCT-guided data-driven framework for tissue shape control, learned from user-guided demonstrations and executed with closed-loop OCT feedback. We study full-volume, tissue-masked sparse-volume, and voxelized point-cloud representations, with or without residual conditioning, using deterministic (Explicit) and generative diffusion (Diffusion) policies.

Across 12 configurations and 576 phantom rollouts on 3-DoF depth and 6-DoF directional tasks, OCT-derived 3D representations support both Explicit and Diffusion policies, with performance depending on representation and residual conditioning. Residual conditioning reduces directional Shape RMSE from 0.748 to 0.555 mm for full-volume Diffusion, while voxelized point-cloud Explicit with residuals achieves 0.534 mm for depth and 0.443 mm for directional control, with comparable accuracy at intermediate unseen targets and consistent performance across deployments. We further evaluate the framework in 60 ex vivo porcine rollouts, extending to biological tissue with final geometric errors comparable to human demonstrations. These results establish volumetric OCT as current and goal representations for learned closed-loop tissue shape control across deterministic and generative action models.

Method overview

OCT acquisition, target generation, three input representations and closed-loop robot policy
Three OCT observations, their relative tool poses, and a fixed target condition the policy. Explicit and Diffusion models predict an eight-action chunk; only the first action is executed before rescanning. RGB provides visual context and is not a policy input.

Three OCT representations

FULL VOLUME

Loading recorded OCT geometry…

Drag to rotate · click, then scroll to zoom

Closed-loop tissue deformation

Phantom depth OCT deformation with a fixed target overlay and RGB contact inset

Closed-loop phantom deformation toward different target depths. Blue: observation; salmon: target. RGB shows the same recorded step.

Numerical results

Twelve configurations are evaluated in 576 closed-loop rollouts over three deployment sweeps. Lower Shape RMSE indicates closer observed and target tissue geometry.

View numerical comparison
Final Shape RMSE (mm), mean ± SD across three target-averaged sweeps. R denotes residual conditioning.
Representation Policy Depth Directional

Residual conditioning lowers final error in nine of twelve matched comparisons.

Citation

BibTeX 
@misc{zhang2026octshapecontrol,
  title = {Robotic Optical Coherence Tomography-Guided Geometric Modeling
           of Soft Tissue: 3D Representations for Data-Driven
           Closed-Loop Shape Control},
  author = {Zhang, Chaoyu and Ma, Guangshen and Draelos, Mark},
  year = {2026},
  note = {Manuscript}
}