Francesca Fardo

This project addresses the challenge of assessing chronic pain, a complex condition affecting around 20% of people. Traditional assessment methods, based on self-reports, fail to capture the multifaceted nature of chronic pain. Our approach utilizes Large Language Models and machine learning to analyze Danish electronic health records, particularly healthcare professionals' notes. These notes are believed to contain linguistic patterns indicative of chronic pain. By applying machine learning to these notes, our goal is to develop a model that can predict individuals at high risk of chronic pain, thus aiding in early, personalized intervention strategies and potentially preventing chronic pain onset.