Deep Learning on Electronic Health Records: Unlocking the Power of Clinical Data - Lecture 1

Explore a comprehensive lecture on leveraging deep learning and natural language processing techniques to unlock the potential of electronic health record (EHR) data for clinical research. Learn about the Community Care Cohort Project (C3PO) and its goal of collecting multimodal EHR data from the Mass General Brigham healthcare system. Discover how pretrained transformer-based models can extract valuable information from clinical text and identify outcome diagnoses from discharge summaries. Gain insights into the architecture and training of large language models (LLMs) like BERT and GPT3, and understand the need for domain-specific models such as BioBERT and ClinicalBERT. Delve into the challenges and ethical considerations of applying LLMs to clinical and biomedical data, including bias and privacy concerns. Finally, learn how to train an LLM on a specific task using the Hugging Face Transformers library.

- Primer
- Meeting