Treatment recommendation for Low Back Pain based on Large Language Models

In the developed world, Low Back Pain (LBP) has a considerable societal 
and economic impact. Patients, in particular chronic patients, experience 
a large descrease in productivity, as well as mental and physical 
wellbeing. In this work, we aim to provide a framework for a decision 
support system for practitioners, based on a dataset provided by the 
University Medical Center Groningen (UMCG). We leverage a combination of 
modern transformer architectures and traditional machine learning models 
to improve f1-score by over 15% (exact result pending) on research 
performed on the same dataset. In addition, we perform a thorough analysis 
on model explainability.