Co-occurring diseases heavily influence the performance of weakly supervised learning models for classification of chest CT
In person: 23 February 2022 • 5:30 PM - 7:00 PM PST
Despite the potential of weakly supervised learning to automatically annotate massive amounts of data, little is known about its limitations for use in computer-aided diagnosis (CAD). For CT specifically, interpreting the performance of CAD algorithms can be challenging given the large number of co-occurring diseases. This paper examines the effect of co-occurring diseases when training classification models by weakly supervised learning, specifically by comparing multi-label and multiple binary classifiers using the same training data. Our results demonstrated that the binary model outperformed the multi-label classification in every disease category in terms of AUC. However, this performance was heavily influenced by co-occurring diseases in the binary model, suggesting it did not always learn the correct appearance of the specific disease.
Duke Univ. (United States), Duke Univ. School of Medicine (United States)
Mr. Fakrul Islam Tushar is a first-year Ph.D. student at the Department of Electrical & Computer Engineering, Pratt School of Engineering, Duke University. Currently, he is a Research Assistant at the Center for Virtual Imaging Trails (CVIT), Duke University, Durham, NC. Primary research focused on building reliable, generalizable, and explainable AI-based solutions for CVIT.