lingshu-medical-mllm/Lingshu-7B
Lingshu-7B is a 7 billion parameter multimodal large language model developed by lingshu-medical-mllm, specifically designed for medical understanding and reasoning. It excels in medical visual question answering (VQA) and report generation, supporting over 12 medical imaging modalities including X-Ray, CT, and MRI. This model achieves state-of-the-art performance on various medical multimodal and textual QA tasks, outperforming many open-source and some proprietary models in its category.
Loading preview...
Lingshu-7B: A Specialist Multimodal Medical LLM
Lingshu-7B is a 7 billion parameter multimodal large language model (MLLM) developed by lingshu-medical-mllm, specifically engineered for advanced medical understanding and reasoning. This model demonstrates state-of-the-art performance across a wide range of medical visual question answering (VQA) and report generation tasks, as detailed in its technical report. It is distinguished by its broad support for over 12 medical imaging modalities, including X-Ray, CT Scan, MRI, Microscopy, Ultrasound, and Histopathology.
Key Capabilities
- State-of-the-Art Medical VQA: Achieves leading performance on benchmarks like SLAKE (83.1), PathVQA (61.9), PMC-VQA (56.3), and OmniMedVQA (82.9) among open-source models under 10B parameters.
- Superior Medical Report Generation: Demonstrates strong results in generating medical reports, particularly on MIMIC-CXR (ROUGE-L 30.8, CIDEr 109.4) and IU-Xray (ROUGE-L 41.2, CIDEr 180.7).
- Extensive Modality Support: Capable of processing and interpreting images from diverse medical sources, enhancing its utility in various clinical applications.
- Competitive Textual QA: Performs robustly on medical textual QA tasks, scoring 74.5 on MMLU-Med and 76.6 on PubMedQA.
Good for
- Medical Image Analysis: Ideal for applications requiring interpretation and question answering based on a wide array of medical imaging data.
- Automated Medical Report Generation: Suitable for assisting in the creation of detailed and accurate medical reports from visual inputs.
- Clinical Decision Support Systems: Can be integrated into systems that require advanced medical reasoning and understanding to aid healthcare professionals.
- Medical Research and Development: A valuable tool for researchers exploring multimodal AI applications in the medical domain.