We are a lab based inside CHIP (Computational Health Informatics Program) at Boston Children’s Hospital, studying the use of natural language processing for biomedical and health-related use cases. Our lab consists of faculty, postdoctoral research fellows, full-time research staff, and affiliated students. While we are primarily technology experts and not medical experts, we are interested in a few specific domains, including mental health, cardiology, and critical care.
Reducing and predicting unplanned readmission are major unmet needs of psychiatric care. This project will use electronic health record databases and computational linguistic approaches to develop a psychiatric specific temporal relation annotation scheme, create readmission prediction models, and compare against expert human performance. Developing readmission prediction tools could not only be used to help target the delivery of resource-intensive interventions to those patients at greatest risk, but also reduce psychiatric health-care costs. NIH RePORTER
This project develops methods for extracting universal patient representations from unstructured text in electronic health records. These methods leverage huge amounts of clinical data, recurrent neural network architectures, and novel training techniques to incorporate information at multiple time scales. These methods are evaluated using public datasets to promote reproducibility, and applied to clinical research tasks that extend the knowledge of patients with pulmonary hypertension and autism spectrum disorder at Boston Children’s Hospital. NIH RePORTER
The major goals of this project are to develop and validate automated methods for classifying high-impact congenital heart disease diagnoses and phenotypes fundamental to pathophysiology and prognosis, and to apply these computable phenotypes to create risk models to predict clinical outcomes. NIH RePORTER
This project is a collaboration between our lab, the Harvard-MIT Center for Regulatory Science, and the United States Food and Drug Administration (USFDA). The goal is to leverage NLP and AI methods to improve the regulation of medical devices.