Implementing Large Language Model (LLM) for Enhanced Tumour Phenotypic & Treatment Recommendations from Electronic Medical Records (EMRs)
Source code for the Indicium Conference 2023 research project: "Implementing Large Language Model (LLM) for Enhanced Tumour Phenotypic & Treatment Recommendations from Electronic Medical Records (EMRs)". We use gpt-3.5-turbo
as our LLM and Flutter to build the front-end for the app.
Uses gpt-3.5-turbo
large language model to:
- extract information from plaintext clinical records
- summarize information for cancers and tumours
- output treatment plan for clinical decision support
To have access to the front-end of this project, you should download Flutter and run:
pip install -r requirements.txt
In addition, please set the OPENAI_API_KEY
either in the environment variable or in config.json
. If you do not have an OpenAI account with API key, please visit OpenAI website
macOS(Bash):
export OPENAI_API_KEY=<your_api_key>
Windows(Command Prompt):
set OPENAI_API_KEY=<your_api_key>
To start the app in Flutter and test different models, please run:
python model.py
or:
python model2.py
This project was developed by the following individuals:
- Caroline Serapio
- Mithusan Sivarajah
- Qasim Tahir
- Simran Dhawan
Special thanks to Yunfei Ma to be the mentor for STEM Fellowship research project.
If you have any questions or feedback regarding this research, feel free to reach out to any of the authors or contributors mentioned above. We are actively looking for feedback from industry experts!
We appreciate your interest and support!