M

Medgemma 4b Pt

Developed by unsloth
MedGemma is a medical multimodal model developed based on Gemma 3, focusing on medical text and image understanding and supporting the construction of healthcare AI applications.
Downloads 105
Release Time : 5/21/2025

Model Overview

MedGemma is a series of model variants based on Gemma 3, specifically trained for text and image understanding in the medical field, aiming to accelerate the development of healthcare AI applications.

Model Features

Multimodal processing capability
Supports simultaneous processing of medical image and text information and can understand complex medical content
High-performance
Outperforms the base Gemma model in various medical benchmark tests
Long context support
Supports long context processing of at least 128K tokens
Medical domain optimization
Specifically optimized for medical fields such as radiology, dermatology, and histopathology

Model Capabilities

Medical image analysis
Medical text generation
Medical question answering
Chest X-ray report generation
Dermatology image classification
Histopathology image understanding

Use Cases

Medical imaging analysis
Chest X-ray report generation
Automatically analyze chest X-ray images and generate diagnostic reports
RadGraph F1 score of 29.5, better than the PaliGemma 2 3B model
Dermatology image classification
Identify and classify skin lesion images
Achieved an accuracy of 71.8% on DermMCQA
Medical question answering
Clinical knowledge question answering
Answer medical-related questions
Achieved an accuracy of 64.4% on MedQA
Radiology visual question answering
Answer questions about medical images
F1 score of 62.3 on SlakeVQA
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase