VinDr SpineXR: Spinal Lesions detection

VinDr-SpineXR is a dataset created by VinBigdata to address the challenge of developing a comprehensive system for classifying and localizing multiple spine lesions from X-ray scans due to the lack of large datasets with high-quality images and human annotations. This dataset contains 10,469 images from 5,000 studies with 13 types of abnormalities manually annotated by expert radiologists, providing radiologist’s bounding-box annotations for developing supervised-learning object detection algorithms. It is currently the largest dataset available for this purpose, and can serve as a benchmark dataset to accelerate the development and evaluation of new machine learning models for spinal X-ray interpretation.