Shaoting Zhang, Computer Science Department
With the ever-increasing amount of annotated medical data, large-scale, data-driven methods provide the promise of bridging the semantic gap between images and diagnoses. The goal of my research work is to increase the scale at which decision support systems can be effective for knowledge discovery in potentially massive databases of medical images. In this talk, we will elaborate two important components of our comprehensive framework for robust and scalable analysis: 1) Robust parsing of medical images. The first step in medical image understanding is robust identification and segmentation of the objects of interest. Our proposed methods provide automatic delineation and measurement of both healthy and abnormal cases, which enables the effective extraction and analysis of information within specific regions. 2) Efficient reasoning in medical image databases. We have developed techniques for large-scale content-based medical image retrieval to provide real-time querying for the most relevant and consistent instances (e.g., similar morphological profiles) for decision support. We will also show some preliminary results on histopathological image analysis.