OphNet
收藏资源简介:
OphNet是由蒙纳士大学信息技术学院AIM实验室创建的大型眼科手术视频基准数据集,旨在理解和分析眼科手术流程。该数据集包含2,278个视频,覆盖66种不同类型的白内障、青光眼和角膜手术,总计约205小时,是现有最大数据集的20倍。OphNet通过精细的层次化标注,包括102种独特的手术阶段和150种操作,支持全面的手术流程理解和解释。数据集的创建过程经过精心设计,由经验丰富的眼科医生进行标注,确保了数据的专业性和准确性。OphNet的应用领域广泛,包括手术流程分析、手术机器人和AI辅助手术系统的开发,以及手术教育和培训,旨在提高手术效率和安全性。
OphNet is a large-scale benchmark dataset of ophthalmic surgical videos developed by the AIM Lab within the School of Information Technology at Monash University, designed to understand and analyze ophthalmic surgical workflows. This dataset contains 2,278 videos covering 66 distinct types of cataract, glaucoma, and corneal surgeries, with a total duration of approximately 205 hours, and it is 20 times the size of the largest existing comparable dataset. OphNet features fine-grained hierarchical annotations, including 102 unique surgical phases and 150 surgical maneuvers, which enable comprehensive understanding and interpretation of surgical workflows. The dataset was developed through a rigorously designed process, with annotations performed by experienced ophthalmologists, ensuring the professional rigor and accuracy of the data. OphNet has a wide range of application domains, including surgical workflow analysis, the development of surgical robots and AI-assisted surgical systems, as well as surgical education and training, with the ultimate goal of enhancing surgical efficiency and safety.




