Development of a System for Determining Technique Level of Vascular Anastomosis Using Hand Motion

• Published in: Journal of robotics and mechatronics Vol. 36; no. 4; pp. 801 - 812
• Main Authors: Chen Xiaoshuai, Shoji Taro, Kowatari Ryosuke, Ebina Koki, Kobayashi Yoshifumi, Kato Moeki, Kabasawa Hinaha, Suzuki Taisei, Sase Kazuya, Tsujita Teppei, Komizunai Shunsuke, Oka Kazuhiko, Konno Atsushi
• Format: Journal Article
• Language: English
• Published: Tokyo Fuji Technology Press Co. Ltd 01.08.2024
• Subjects: Classification; Evaluation; Machine learning; Surgeons
• ISSN: 0915-3942; 1883-8049
• DOI: 10.20965/jrm.2024.p0801

In off-pump coronary artery bypass grafting (OPCAB), the coronary arteries are joined as the heart beats. This procedure requires high skill and experience to be performed reliably and quickly. Although training kits are commonly used for technical training, the inability of trainees to always be with experienced surgeons for guidance and to receive immediate evaluation remains problematic. To address this problem, a system that allows a single trainee to observe and quantitatively evaluate the procedures performed by an experienced surgeon is being developed. In this study, to analyze the differences between the motions of experienced and novice surgeons, Leap Motion was used to measure the hand motion of the vascular anastomosis performed by both surgeons using a training kit. Using the measured data, the features of the vascular anastomosis surgical techniques performed by experienced and novices were tested using the Mann–Whitney U test. In addition, a binary classification was performed using machine learning with these features. The binary classification results show that trainees can be classified as experts or novices with high accuracy using the developed system.