Speaker
The electronic health record is a rich and still largely untapped source of clinical information that can be coupled with machine learning to support and augment clinical decision-making. However, much of this information is present in clinical notes, images, and other unstructured formats rather than in structured fields. In this 1-hour seminar, we'll discuss why unstructured data is so valuable in many prediction tasks, then learn how to incorporate features from multiple modalities in a single, neural network based predictive model. Although our discussion will focus primarily on electronic health record data, the proposed approach can be applied to a wide range of multi-modal data sources. This event is being cross-promoted by the NC BERD Consortium, a collaboration of the CTSA-funded BERD cores at UNC-Chapel Hill, Wake Forest University School of Medicine, and Duke University School of Medicine.