In this talk we describe a new set of computational tools for manipulating pixel intensities based on optimal transport and hierarchical learning. We will show how these methods can be used to uncover predictive information that is too complex to be extracted by visual examination of raw image data. We will demonstrate these concepts in tasks related to drug discovery, diagnosis of cytology and histopathology images, acinar tissue quantification in pancreatitis from histology, amongst other applications.
Healthcare institutions are now recording more electronic health data about patients than ever before, including data about patient conditions, lab tests, genomics, treatments, and outcomes. However, an open question remains on what one can do with all of this data. Many hope that if researchers tap into this real world observational data, the collective experience of the healthcare system can be leveraged to unearth insights to improve the quality of care.
Biological processes, including those involved in immune response and disease progression, are often dynamic. To model the regulatory and signaling networks that are activated as part of these systems we are developing methods to combine the abundant static regulatory, proteomic and epigenetic data with time series gene and miRNA expression data. The reconstructed networks characterize the pathways involved in the response, their time of activation, and the affected genes.
One of the most fundamental and fascinating questions in biology is how different types of cells in a human body are derived from the same genome but possess distinct appearances and functions. Temporal and spatial- specific gene transcription, which is tightly controlled by cis-regulatory elements such as promoters and enhancers, has been regarded as one of the main contributor.
Andrew King - Electronic medical records (EMRs) are capturing increasing amounts of data per patient. For clinicians to efficiently and accurately understand a patient’s clinical state, better ways are needed to determine when and how to display EMR data.
The PaTH Network is a DBMI-led clinical data research network (CDRN) within the Patient Centered Outcomes Research Institute (PCORI). This project is building a network of EMR data from six institutions (Temple University, UPMC, Penn State University, Johns Hopkins University, University of Utah, and Geisinger Health System) covering roughly 11 million patients.