Force-regulated robotic diffuse reflectance spectroscopy scanning
File(s)
Accepted version
Author(s)
Type
Journal Article
Abstract
Significance: Intraoperative discrimination between malignant and benign tissue is essential for achieving complete tumour resection while preserving healthy structures. Robotic diffuse reflectance spectroscopy (DRS) enables automated intraoperative tissue assessment, but unregulated probe–tissue contact forces introduce spectral artifacts and risk tissue damage, limiting clinical translation.
Aim: Wepresent and evaluate an autonomous robotic DRS scanning system incorporating real-time contact forceregulation to enable safe, repeatable, and large-area ex vivo tissue assessment.
Approach: A serial robot manipulator equipped with a fibre-optic DRS probe, load cell, and camera employed
hybrid visual servoing and force feedback control, maintaining 0.08N target force with < 0.30N peak limits. Performance was evaluated on ex vivo bovine and ovine tissues including scanning safety, force regulation versus expert operators, spectral consistency across six novice operators, and tissue classification using support vector machines.
Results: The system achieved 100% and 93% scan completion rates on bovine and ovine tissues, respectively, without visible tissue damage. Autonomous operation enabled 46 continuous trials over approximately 30 minutes without human intervention. Mean applied force (0.079N) matched manual sampling (0.080N) with reduced variance
and lower peak forces. Robotic spectral acquisitions aligned with manual measurements, yielding mean inter-observer spectral angles of 1.43◦, comparable to inter-human operator variability. Tissue classification demonstrated high discriminatory accuracy in bench-top validation.
Conclusions: Force-regulated robotic DRS enables standardised, autonomous tissue scanning with expert-level spectral consistency, advancing spectroscopic sensing toward reliable automated surgical margin assessment.
Aim: Wepresent and evaluate an autonomous robotic DRS scanning system incorporating real-time contact forceregulation to enable safe, repeatable, and large-area ex vivo tissue assessment.
Approach: A serial robot manipulator equipped with a fibre-optic DRS probe, load cell, and camera employed
hybrid visual servoing and force feedback control, maintaining 0.08N target force with < 0.30N peak limits. Performance was evaluated on ex vivo bovine and ovine tissues including scanning safety, force regulation versus expert operators, spectral consistency across six novice operators, and tissue classification using support vector machines.
Results: The system achieved 100% and 93% scan completion rates on bovine and ovine tissues, respectively, without visible tissue damage. Autonomous operation enabled 46 continuous trials over approximately 30 minutes without human intervention. Mean applied force (0.079N) matched manual sampling (0.080N) with reduced variance
and lower peak forces. Robotic spectral acquisitions aligned with manual measurements, yielding mean inter-observer spectral angles of 1.43◦, comparable to inter-human operator variability. Tissue classification demonstrated high discriminatory accuracy in bench-top validation.
Conclusions: Force-regulated robotic DRS enables standardised, autonomous tissue scanning with expert-level spectral consistency, advancing spectroscopic sensing toward reliable automated surgical margin assessment.
Date Acceptance
2026-09-08
Citation
Journal of Biomedical Optics
ISSN
1083-3668
Publisher
SPIE
Journal / Book Title
Journal of Biomedical Optics
Copyright Statement
Copyright This paper is embargoed until publication. Once published the Version of Record (VoR) will be available on immediate open access.
License URL
Publication Status
Accepted