LMOD
收藏资源简介:
LMOD是一个大型多模态眼科数据集,由耶鲁大学等机构创建,旨在评估大型视觉-语言模型在眼科图像上的表现。数据集包含21,993张图像,涵盖光学相干断层扫描、扫描激光眼底镜、眼照片、手术场景和彩色眼底照片等多种眼科成像方式。数据集创建过程中,对图像进行了多粒度标注,包括边界框、区域标注和图像级标注。LMOD的应用领域主要集中在眼科疾病的诊断和分类,旨在通过AI技术提高眼科临床工作流程的效率和准确性。
LMOD is a large-scale multimodal ophthalmology dataset developed by institutions including Yale University, aiming to evaluate the performance of large vision-language models on ophthalmic images. The dataset contains 21,993 images covering multiple ophthalmic imaging modalities, including Optical Coherence Tomography (OCT), Scanning Laser Ophthalmoscope (SLO), ophthalmic photographs, surgical scenes, and color fundus photographs. During its development, the dataset underwent multi-granularity annotations covering bounding boxes, regional annotations, and image-level annotations. Its primary application focuses on the diagnosis and classification of ophthalmic diseases, with the goal of enhancing the efficiency and accuracy of ophthalmic clinical workflows through AI technologies.




