TongueAtlas-4K: A Large-Scale Tongue Imaging Benchmark for AI Diagnosis

Description TongueAtlas-4K is a large-scale dataset consisting of 4,000 tongue images collected from clinical patients, annotated by certified Traditional Chinese Medicine (TCM) experts. This dataset is designed to support research in AI-assisted tongue diagnosis, including tasks such as image classification, syndrome recognition. This dataset is associated with the paper "MIRNet: Integrating Constrained Graph-Based Reasoning with Pre-training for Diagnostic Medical Imaging" (AAAI 2026, to appear). If you use this dataset, please cite the paper. Dataset Contents- `images.zip`: Contains 4,000 raw tongue image files in JPG format.- `labels.xlsx`: Contains image filenames and their corresponding labels for 22 tongue appearance categories.