白内障手术数据集
收藏资源简介:
该数据集由伦敦帝国理工学院的研究团队创建,主要用于白内障手术视频中的手术器械识别。数据集包含615张图像,涵盖了10种不同的手术器械类别,如套管、新月刀、固定环等。数据集的创建过程包括从公开的教学视频中逐帧提取图像,并通过Roboflow进行标注。数据集经过图像增强处理,以提高数据质量和数量。该数据集的应用领域主要集中在医学教育和手术辅助系统,旨在通过深度学习模型实时识别手术器械,帮助眼科医生和学员更好地理解手术过程。
This dataset was developed by a research team at Imperial College London, primarily for surgical instrument recognition in cataract surgery videos. It contains 615 images covering 10 distinct surgical instrument categories, including cannulas, sickle knives, retention rings, and others. The dataset construction workflow involves extracting images frame by frame from publicly accessible educational surgical videos and performing annotations via Roboflow. Image augmentation has been applied to the dataset to enhance both its overall quality and scale. The main application areas of this dataset are medical education and surgical assistance systems. Its core goal is to enable real-time surgical instrument recognition using deep learning models, thereby assisting ophthalmologists and trainees in gaining a better understanding of the surgical process.

- 1Identifying Surgical Instruments in Pedagogical Cataract Surgery Videos through an Optimized Aggregation Network伦敦帝国理工学院 · 2025年



