Artificial Intelligence Driven Tools for Objective Surgical Performance Improvement

摘要/摘要 目前，受监督的外科培训只提供了执业生涯中外科经验的一小部分。 外科医生。外科医生的技能在他们的职业生涯中不断发展。外科医生从专家的监督反馈中受益 但一旦他们开始独立实践，他们就会失去这种结构化和具体的反馈。手术技巧是 与患者的预后有关。因此，通过自动化支持外科医生的持续专业学习 结构化资源可以改善患者护理。外科医生在实践中的现状是衡量患者的预后或 其他护理过程变量作为他们技能的间接衡量标准。这些措施并没有告诉外科医生如何改善。 这个项目的目标是开发工具来分析外科领域的视频，为外科医生提供公正的技能 关于如何改进的评估和具体反馈。该项目包括将这些工具集成到个性化的 外科医生技能获取的外科学习平台及其效果评价。为了实现这个目标，这个项目 包括一个多学科团队，包括眼科、外科教育、外科数据科学、计算机 视觉、机器学习和深度学习、统计学和人机交互。开发的视频分析工具 这个项目将使以下白内障手术成为可能，这是美国最常见的外科手术之一， 在世界各地：1)对外科医生的技能进行客观评估；2)为外科医生提供关于如何改进的具体反馈 考虑到他们过去的表现，这是个性化的；以及3)个性化学习有效性的初步证据 外科医生技能获取的平台。我们工作的预期影响是创造一条途径，在其中外科医生 激励他们将自己和他们的表现视为改善护理成果和价值的过程的一部分，以及 机构可以使用客观工具来制定可重复使用的外科手术能力标准。

Abstract / Summary Currently, supervised surgical training provides only a small fraction of surgical experience in the career of a practicing surgeon. Surgeons’ skill develops throughout their career. Surgeons benefit from supervised feedback from experts during training, but they lose such structured and specific feedback once they begin independent practice. Surgical skill is associated with patient outcomes. Therefore, supporting surgeons’ continuous professional learning through automated structured resources can improve patient care. The status quo for surgeons in practice is to measure patient outcomes or other process of care variables as indirect measures of their skill. These measures do not inform surgeons how to improve. The goal in this project is to develop tools to analyze videos of the surgical field to provide surgeons with unbiased skill assessments and specific feedback on how to improve. This project includes integration of these tools into a personalized surgical learning platform and evaluation of its effectiveness for surgeons’ skill acquisition. To achieve this goal, this project includes a multi-disciplinary team to include expertise in ophthalmology, surgical education, surgical data science, computer vision, machine learning and deep learning, statistics, and human-computer interaction. The video analysis tools developed in this project will enable the following for cataract surgery, one of the most common surgical procedures in the U.S. and across the world: 1) objective assessments of surgeons’ skill; 2) provide surgeons with specific feedback on how to improve that is personalized given their past performance; and 3) preliminary evidence of effectiveness of a personalized learning platform for surgeons’ skill acquisition. The anticipated impact of our work is to create a pathway in which the surgeon is incentivized to see themselves and their performance as part of the process of improving outcomes and value in care, and institutions have access to objective tools to create reproducible standards for surgical competency.