Applications of Deep Learning in Healthcare

My group is invested in the development of machine learning methodology with applications in health and disease. Currently we are mostly interested in leveraging representation learning and deep learning architectures to build models to represent data from different modalities such as longitudinal structured data (e.g., electronic health records), images (e.g., tomography, ultrasound) and natural language (clinical narrative), to be used in predictive tasks of relevance in healthcare. For instance, classification or estimation of risk in diverse clinical outcomes, and to address challenges associated with performance characterization in predictive models.