Paper
Force-based Automatic Classification of Basic Manipulations with Grasping Forceps
- Authors:
- Yuichi Kurita; Toshio Tsuji; Tomohiro Kawahara; Masazumi Okajima; Hiroyuki Egi; Hideki Ohdan; Tsukasa Ogasawara
- Abstract
- Haptic information is crucial for the execution of precise and dexterous manipulations. During minimally invasive surgery, doctors are required to indirectly sense force-related information from body organs and tissues via a surgical instrument because they cannot directly touch the tissue. Against such a background, skill evaluation based on force measurement is useful for judging whether a person has adequate manipulation skills. This paper addresses the challenge of automatically classifying basic manipulations performed with surgical grasping forceps. First, manipulations performed with forceps during laparoscopic surgery were categorized into four basic types from video observation. Grasping forceps with force-sensing capability were developed to support identification of these types, which were automatically classified by monitoring information on the force applied to the forceps. An experiment to investigate the efficacy of the proposed method produced manipulation logs showing that doctors are capable of conducting tasks with less force than novices. It was also confirmed that the prototype forceps are suitable for practical use in animal experiments.
- Keywords
- Minimally Invasive Surgery; Forceps Manipulation; Automatic Classification
- StartPage
- 76
- EndPage
- 82
- Doi
- 10.5963/LSMR0302005