Abu Dhabi-based company M42 Health has released Med42, an open-access clinical large language model (LLM) designed to enhance public access to advanced AI capabilities in healthcare. Med42, built using a human-curated medical literature and patient information dataset, outperforms existing medical AI models and offers accurate responses to medical inquiries. While further validation is needed, Med42 has the potential to improve clinical decision-making and expand access to medical knowledge.
M42 Introduces Med42: An Open-Access Clinical Large Language Model (LLM) to Expand Access to Medical Knowledge
M42 Health, based in Abu Dhabi, UAE, has just released Med42, an innovative open-access clinical large language model. This model has 70 billion parameters and has the potential to revolutionize healthcare by providing advanced AI capabilities to the public.
Med42, which is based on Meta’s Llama-2 – 70B model, outperforms previous open-source medical AI models by a significant margin. It surpasses OpenAI’s ChatGPT 3.5 in medical question-answering, achieving up to 72% accuracy on the USMLE. This demonstrates its ability to assist doctors in making clinical decisions by providing synthesized medical knowledge.
The Med42 model was developed by the M42 Health AI team using a curated dataset of medical literature and patient information. It was fine-tuned using the Condor Galaxy 1 supercomputer by M42, Cerebras, and Core42. The efficacy of the model was also evaluated by experts at the Mohamed bin Zayed University for Artificial Intelligence (MBZUAI).
Key Benefits of Med42:
- Adaptability: Med42 has the potential to significantly impact medical judgment and can be used for various tasks such as generating personalized treatment plans and analyzing medical material quickly.
- Improved Clinical Decision-Making: Med42 can enhance clinical decision-making by providing accurate responses to health-related queries and supporting medical diagnosis.
- Increased Access to Medical Information: Med42 is a free, publicly available model that aims to make medical information more accessible to the public.
Possible Applications of Med42:
- Answering Health-Related Questions
- Synopsis of Medical History
- In support of medical diagnosis
- Common Health Questions
The code and weights of Med42 have been released to Hugging Face, enabling scientific collaboration and research. The licensing terms are similar to Meta’s Llama 2 model, allowing free research and non-commercial usage while considering the risks associated with using AI in healthcare.
Performance Indicators:
- Med42 achieves 72% accuracy on a sample exam of USMLE, outperforming other publicly available medical LLMs.
- Results on the MedQA dataset show 61.5% accuracy, compared to 50% for GPT-3.5.
- Med42 consistently performs better than GPT-3.5 on MMLU clinical issues.
While the findings are promising, further validation is required before Med42 can be used in clinical practice. It is crucial to address potential inaccuracies, harmful results, and biases in training data. Responsible testing and validation are essential to ensure the safe and effective use of Med42.
Med42 represents a significant advancement in medical AI, emphasizing the importance of ethics and safety in research and development. Its open-access publication allows researchers worldwide to benefit from its capabilities. With thorough validation, models like Med42 have the potential to improve healthcare decision-making globally. Continued openness and collaboration are key to realizing the full potential of Med42.
For more information, visit the Project Page.
All credit for this research goes to the researchers on this project.
If you want to evolve your company with AI and stay competitive, consider using M42’s Med42 to expand access to medical knowledge. To get started, follow these steps:
- Identify Automation Opportunities: Locate key customer interaction points that can benefit from AI.
- Define KPIs: Ensure your AI endeavors have measurable impacts on business outcomes.
- Select an AI Solution: Choose tools that align with your needs and provide customization.
- Implement Gradually: Start with a pilot, gather data, and expand AI usage judiciously.
For AI KPI management advice, reach out to us at hello@itinai.com. Stay updated on the latest AI research news and projects by joining our ML SubReddit, Facebook Community, Discord Channel, and Email Newsletter.
Spotlight on a Practical AI Solution: AI Sales Bot
Consider using the AI Sales Bot from itinai.com/aisalesbot to automate customer engagement and manage interactions across all stages of the customer journey. This solution can redefine your sales processes and improve customer engagement.
Explore AI solutions at itinai.com.
