Add dataset card and link to paper

README.md

@@ -1,3 +1,37 @@
 ---
 license: apache-2.0
+task_categories:
+- image-text-to-text
+language:
+- zh
+tags:
+- medical
 ---
+
+# MedRealMM: A Real-World Multimodal Benchmark for Chinese Online Medical Consultation
+
+This repository contains the dataset for **MedRealMM**, presented in the paper [MedRealMM: A Real-World Multimodal Benchmark for Chinese Online Medical Consultation](https://huggingface.co/papers/2607.09142).
+
+## Dataset Summary
+
+**MedRealMM** is a large-scale, real-world multimodal benchmark for evaluating Large Language Models (LLMs) in Chinese online medical consultation. Built from de-identified patient-doctor interactions collected from a nationwide Chinese internet hospital, MedRealMM preserves the authentic text-image contexts that physicians face in actual clinical practice.
+
+The dataset contains **5,620 real-world multimodal cases** spanning 64 clinical departments.
+
+## Dataset Features
+
+- **Multimodal Context**: Includes both text-based consultation history and patient-uploaded medical images.
+- **MCCP Framework**: Uses a Multimodal Clinical Challenge Point (MCCP) extraction framework to identify clinically demanding moments in consultation trajectories and converts them into standardized next-response generation tasks.
+- **Clinician-Refined Rubrics**: Each instance is paired with a case-specific rubric refined by physicians to evaluate open-ended responses based on clinically desirable and undesirable behaviors.
+
+## Citation
+
+If you use MedRealMM in your research, please cite:
+
+```bibtex
+@article{medrealmm2026,
+  title={MedRealMM: A Real-World Multimodal Benchmark for Chinese Online Medical Consultation},
+  journal={arXiv preprint arXiv:2607.09142},
+  year={2026}
+}
+```