3D Vision Based Robot Assisted Electrical Impedance Scanning for Soft Tissue Conductivity Sensing
| Autor: | Marco Piccinelli, Zhuoqi Cheng, Diego Dall'Alba, Michael Kjaer Schmidt, Thiusius Rajeeth Savarimuthu, Paolo Fiorini |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2022 |
| Předmět: |
Conductivity
Mathematical models Control and Optimization tissue identification Sensors Mechanical Engineering Biomedical Engineering finite element modeling vision based sensing Surgical robotics Computer Science Applications Human-Computer Interaction Electric potential Artificial Intelligence Control and Systems Engineering Robot sensing systems Probes Computer Vision and Pattern Recognition Electrodes electrical bio-impedance sensing |
| Zdroj: | Piccinelli, M, Cheng, Z, Dall'Alba, D, Schmidt, M K, Savarimuthu, T R & Fiorini, P 2022, ' 3D Vision Based Robot Assisted Electrical Impedance Scanning for Soft Tissue Conductivity Sensing ', IEEE Robotics and Automation Letters (RA-L), vol. 7, no. 2, pp. 4055-4062 . https://doi.org/10.1109/LRA.2022.3150481 |
| DOI: | 10.1109/LRA.2022.3150481 |
| Popis: | Advanced Sensing Technologies (ASTs) have a great potential to improve surgical quality and to further develop Surgical Robotic Systems (SRSs), enhancing their technical and autonomy capabilities. Among these sensing techniques, Electrical Bioimpedance (EBI) provides a non-invasive, low-cost, and safe AST for the intraoperative localization of abnormal regions. The current EBI integration into SRS has only been demonstrated in an over-simplified condition (i.e. nearly flat surfaces), which are almost never encountered in real anatomies. To overcome this limitation, we develop a robotic assisted EBI scanning system able to work with tissues' surfaces of arbitrary shapes, leveraging 3D vision based tissue reconstruction in the scanning process. In addition, we propose a novel model based conductivity estimation method that exploits Finite Element (FE) simulation to compensate for errors introduced by non-planar surfaces and uncertainty in the electrodes' position. The system is evaluated through experiments in simulation and using ex vivo animal tissues. The experimental results show that the model based method achieves an accuracy of 99% independently of the curvature of the tissue surface, while the previous method achieves an accuracy ranging from 70% to 88% depending on the surface curvature. The obtained results are very promising and show a great potential to be integrated into existing SRSs for identifying different tissues during a robotic surgery without involving any additional tool. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|