Skip to content

Latest commit

 

History

History
16 lines (13 loc) · 976 Bytes

README.md

File metadata and controls

16 lines (13 loc) · 976 Bytes

synthetic-clinical-notes

Create and analyse synthetic clinical notes

Goal

Experiment with generating and analysing synthetic medical notes using Large Language Models (LLMs).

Objectives

  1. Learn to create a Reproducible Analytical Pipeline (RAP). Learn More
  2. Develop prompt engineering skills for LLMs.
  3. Create a dataset of synthetic medical notes for general use.
  4. Learn how to apply lightweight LLMs (such as could be implemented on premise for a hospital) to clinical notes.

Why Synthetic Medical Notes?

  1. There are few datasets that allow a researcher to experiment with LLMs (like ChatGPT, Claude, open source options or other models to come on) medical notes.
  2. Synthetic notes allow safe experimentation without exposing any patient identifying data.
  3. Doing this project will enable me to assess quality of synthetic text for my own research.