Towards Motion Correction in Robotized Medical Procedures Using Real-Time 3D Body Measurement - 25.10
Title:
Towards Motion Correction in Robotized Medical Procedures Using Real-Time 3D Body Measurement
Authors:
Matej CAMPELJ, Matija JEZERSEK
Faculty of Mechanical Engineering, University of Ljubljana, Ljubljana, Slovenia
Full paper:
Keywords:
3D body scanning, optical flow, point cloud, TOF camera
Abstract:
Robotized medical procedures rely on high spatial accuracy and repeatability, yet patient motion during treatment remains a critical challenge that can compromise safety and effectiveness. In particular, procedures such as laser skin rejuvenation require continuous adaptation of the robot trajectory to unintentional movements of the patient. To address this, we propose a real-time motion correction framework that combines dense optical flow with 3D point cloud data acquired by a time-of-flight camera.
The approach, termed Fusion Flow, transforms pixel-level motion fields into real-world displacements
in all three spatial directions and integrates them into the global trajectory of the treatment area.
We evaluated seven optical flow algorithms (DeepFlow, DenseRLOF, DualTVL1, Farneback, PCAFlow,
SimpleFlow, and SparseToDense) using a controlled experimental setup with sub-millimeter reference
motion. Among them, the Farneback method achieved the best balance of speed and accuracy,
operating at 140 FPS with an average trajectory error of only 8 mm over a 2250 mm path.
These results demonstrate that real-time fusion of optical flow and 3D depth data provides clinically relevant precision for motion correction, paving the way for safer and more effective robotized medical procedures.
DOI:
https://doi.org/10.15221/25.10
How to Cite (MLA):
Campelj, Matej and Jezersek, Matija, "Towards Motion Correction in Robotized Medical Procedures Using Real-Time 3D Body Measurement", 3DBODY.TECH Journal, vol. 2, Dec. 2025, #10, https://doi.org/10.15221/25.10.
Presentation:
Video availble in 3DBODY.TECH 2025 Proceedings - 25.10
Details:
Volume/Issue: 3DBODY.TECH Journal - Vol. 2, 2025
Paper: #10
Published: 2025/12/31
Presented at: 3DBODY.TECH 2025, 21-22 Oct. 2025, Lugano, Switzerland
Proceedings: 3DBODY.TECH 2025 Proceedings
License/Copyright notice
Copyright © 2025 by the author(s).
This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
The papers appearing in the journal reflect the author's opinions. Their inclusion in the volumes does not necessary constitute endorsement by the editor or by the publisher.
Note: click the + on the top left of the page to open/close the menu.