80% of patient information is typically found within the free text documents of Electronic Health Records (EHR). Traditionally, this information is extracted manually which is time-consuming, inefficient and impractical at scale. This project aims to assess how Natural Language Processing (NLP) can be used to automate the extraction of relevant clinical information from EHR and the explainability of these NLP models.

This project utilises MedCAT - an open source annotation tool to extract oncology concepts/training data and Clinical-BERT to train a NER model.

Please see project summary for more info.