A Python project that creates specialized LLM-based AI agents to analyze complex medical cases.
The system integrates insights from different medical specialists to provide comprehensive assessments
and suggested treatment directions, demonstrating the potential of AI in multidisciplinary medicine.

⚠️ Disclaimer: This project is for research and educational purposes only.
It is not intended for clinical use.

• Added MIT License
• Fixed bugs and updated requirements.txt
• Added .gitignore
• Upgraded core LLM to GPT-5

In the current version, we use three AI agents (GPT-5), each specializing in a different aspect of medical analysis.
A medical report is passed to all agents, which run in parallel (threading) and return their findings.
The outputs are then combined and summarized into three possible health issues with reasoning.

1. Cardiologist Agent

• Focus: Detect cardiac issues such as arrhythmias or structural abnormalities.
• Recommendations: Cardiovascular testing, monitoring, and management strategies.

2. Psychologist Agent

• Focus: Identify psychological conditions (e.g., panic disorder, anxiety).
• Recommendations: Therapy, stress management, or medication adjustments.

3. Pulmonologist Agent

• Focus: Assess respiratory causes for symptoms (e.g., asthma, breathing disorders).
• Recommendations: Lung function tests, breathing exercises, respiratory treatments.

• Medical Reports/ → Synthetic medical report samples
• Results/ → Outputs generated by the agents

1. Clone the repo:

   git clone https://github.com/ahmadvh/AI-Agents-for-Medical-Diagnostics.git
   cd AI-Agents-for-Medical-Diagnostics

2. Create a virtual environment and install dependencies:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   pip install -r requirements.txt

3. Set up your API credentials:
   • Create a file named apikey.env in the project root.
   • Add your OpenAI (or other LLM provider) credentials:

   OPENAI_API_KEY=your_api_key_here

4. Run the system: python main.py

Planned improvements for upcoming versions include:

• Specialist Expansion: Add new agents for Neurology, Endocrinology, Immunology, and other fields.
• Local LLM Support: Integrate models such as Llama 4 via Ollama, vLLM, or llama.cpp, with function-calling style hooks and safe code execution.
• Vision Capabilities: Enable multimodal decision-making with agents that analyze radiology images and other medical scans.
• Live Data Tools: Incorporate LLM-based tools for real-time search and querying structured medical datasets.
• Advanced Parsing: Improve handling of complex medical reports with structured outputs (e.g., JSON schema validation).
• Automated Testing: Add evaluation pipelines and smoke-test CI with mocked LLM calls for reproducibility.

This repository is licensed under the MIT License.

You are free to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the software, subject to the conditions described in the LICENSE file.

The software is provided “as is”, without warranty of any kind.