Alejandro Zulbaran-Rojas1, Bijan Najafi1, Nestor Arita1, Hadi Rahemi1, Javad Razjouyan1, Ramyar Gilani2. 1. Division of Vascular Surgery and Endovascular Therapy, Michal E DeBakey Department of Surgery, Baylor College of Medicine, Houston, Texas. 2. Division of Vascular Surgery and Endovascular Therapy, Michal E DeBakey Department of Surgery, Baylor College of Medicine, Houston, Texas. Electronic address: rgilani@bcm.edu.
Abstract
BACKGROUND: Traditional assessment (e.g., checklists, videotaping) for surgical proficiency may lead to subjectivity and does not predict performance in the clinical setting. Hand motion analysis is evolving as an objective tool for grading technical dexterity; however, most devices accompany with technical limitations or discomfort. We purpose the use of flexible wearable sensors to evaluate the kinematics of surgical proficiency. METHODS: Surgeons were recruited and performed a vascular anastomosis task in a single institution. A modified objective structured assessment of technical skills (mOSATS) was used for technical qualification. Flexible wearable sensors (BioStamp RCTM, mc10 Inc., Lexington, MA) were placed on the dorsum of the dominant hand (DH) and nondominant hand (nDH) to measure kinematic parameters: path length (Tpath), mean (Vmean) and peak (Vpeak) velocity, number of hand movements (Nmove), ratio of DH to nDH movements (rMov), and time of task (tTask) and further compared with the mOSATS score. RESULTS: Participants were categorized as experts (n = 12) and novices (n = 8) based on a cutoff mean mOSATS score. Significant differences for tTask (P = 0.02), rMov (P = 0.07), DH Tpath (P = 0.04), Vmean (P = 0.07), Vpeak (P = 0.04), and nDH Nmove (P = 0.02) were in favor of the experts. Overall, mOSATS had significant correlation with tTask (r = -0.69, P = 0.001), Nmove of DH (r = -0.44, P = 0.047) and nDH (r = -0.66, P = 0.001), and rMov (r = 0.52, P = 0.017). CONCLUSIONS: Hand motion analysis evaluated by flexible wearable sensors is feasible and informative. Experts utilize coordinated two-handed motion, whereas novices perform one-handed tasks in a hastily jerky manner. These tendencies create opportunity for improvement in surgical proficiency among trainees.
BACKGROUND: Traditional assessment (e.g., checklists, videotaping) for surgical proficiency may lead to subjectivity and does not predict performance in the clinical setting. Hand motion analysis is evolving as an objective tool for grading technical dexterity; however, most devices accompany with technical limitations or discomfort. We purpose the use of flexible wearable sensors to evaluate the kinematics of surgical proficiency. METHODS: Surgeons were recruited and performed a vascular anastomosis task in a single institution. A modified objective structured assessment of technical skills (mOSATS) was used for technical qualification. Flexible wearable sensors (BioStamp RCTM, mc10 Inc., Lexington, MA) were placed on the dorsum of the dominant hand (DH) and nondominant hand (nDH) to measure kinematic parameters: path length (Tpath), mean (Vmean) and peak (Vpeak) velocity, number of hand movements (Nmove), ratio of DH to nDH movements (rMov), and time of task (tTask) and further compared with the mOSATS score. RESULTS: Participants were categorized as experts (n = 12) and novices (n = 8) based on a cutoff mean mOSATS score. Significant differences for tTask (P = 0.02), rMov (P = 0.07), DH Tpath (P = 0.04), Vmean (P = 0.07), Vpeak (P = 0.04), and nDH Nmove (P = 0.02) were in favor of the experts. Overall, mOSATS had significant correlation with tTask (r = -0.69, P = 0.001), Nmove of DH (r = -0.44, P = 0.047) and nDH (r = -0.66, P = 0.001), and rMov (r = 0.52, P = 0.017). CONCLUSIONS: Hand motion analysis evaluated by flexible wearable sensors is feasible and informative. Experts utilize coordinated two-handed motion, whereas novices perform one-handed tasks in a hastily jerky manner. These tendencies create opportunity for improvement in surgical proficiency among trainees.
Authors: Jani Koskinen; Antti Huotarinen; Antti-Pekka Elomaa; Bin Zheng; Roman Bednarik Journal: Int J Comput Assist Radiol Surg Date: 2021-12-15 Impact factor: 2.924