Visão computacional orientando técnica microcirúrgica em simulador ex-vivo

Abstract

1 Abstract 1.1 Background Mastering cerebrovascular microsurgical techniques requires intensive training, with hours spent into a laboratory. Ability learning and maintenance generally demand a supervisor to improve, and one is not always available, specially outside of an operating room setting. Using a software coach and a placenta could help a neurosurgeon in training to improve his abilities. 1.2 Methods: Five volunteer medical doctors with different levels of microsurgical expertize performed vessel dissection and micro-suture simulation in an ex-vivo model. Procedures were video recorded with color labeled instruments for computer analyzes of time, number of right and left hand movements, suture symmetry, vessel stenosis and leakage. Proficiency index software was suggested, in a self-challenge and self-coaching manner. 1.3 Results Proficiency index of microsurgical education (PRIME) pilot results were individually based in an intra-user analysis of performance. No statistical significance was found. 1.4 Conclusion PRIME could offer the capability of constant automated laboratory microsurgical practice under guidance, opening a new window of oriented training without a human tutor in a practical and feasible manner. Prospective, large data study is needed to confirm this hypothesis. Keywords: microsurgery, cerebrovascular surgery, computer vision, artificial intelligence, neurosurgical training, neurosurgical education, laboratory training