Broca – a large language model to draft highly personalized preliminary radiology reports for x-rays

The proposed research project aims to develop a machine learning model that can draft preliminary radiology reports for x-rays. The project will use a large language model based on transformer neural networks to analyze x-rays and generate reports that are personalized to the reporting radiologist and based on the patient’s x-ray images. The goal is to help radiologists be more efficient and reduce the risk of burnout. By automating the repetitive and tedious task of reporting x-rays, radiologists can focus on more complex and important tasks. The end result will be a tool that can improve the quality of healthcare and make it more accessible for all patients, supporting the mission of the partner organization, 16 Bit.

Faculty Supervisor:

Farzad Khalvati

Student:

Partner:

16 Bit

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

University of Toronto

Program: