6D robotic OCT scanning of curved tissue surfaces

Optical coherence tomography (OCT) is a non-invasive volumetric imaging modality with high spatial and temporal resolution. Imaging larger tissue regions requires moving the OCT probe to cover the target area. Handheld scanning approaches rely on overlapping volumes for stitching, whereas, robotic scanning approaches often restrict motion to pure translations to avoid full hand-eye calibration - a challenging problem given the OCT probe's small field of view. These registration- or translation-based stitching strategies, however, struggle when imaging curved tissue surfaces. We propose a fiducial marker enabling full six-degree-of-freedom hand-eye calibration for a robot-mounted OCT probe. Our method produces highly repeatable transform estimates and, in experiments on two phantom surfaces, enables consistent robotic scanning of large, curved geometries. Because the approach does not depend on image-based registration, it avoids error accumulation along scanning paths and outperforms conventional purely translational robotic scanning.

DDC Class

620: Engineering

Funding(s)

Präzise, markerlose semantische Lageverfolgung von Weichgeweben

Centre of Excellence of Al for Sustainable Living and Working

Options

6D robotic OCT scanning of curved tissue surfaces