Patient Similarity Learning through Distance Metric Learning and Interactive Visualization¶
Jimeng Sun
Dr.
IBM TJ Watson Research Center
Abstract:
The objective of patient similarity is to quantitatively measure how similar patients are to each other. The challenges of comprehensive patient similarity are the following:
· How to leverage physician feedback into the similarity computation?
· How to integrate multiple sources of clinical information for patient similarity computation?
· How to compare patients at different stages of disease progression?
· How to incrementally update the existing patient similarity functions as new data arrive?
· How to present the similarity in an intuitive way?
In this work, we will present the comprehensive patient similarity framework that answers those questions. The core of the framework is the combination of advanced distance metric learning algorithms and novel visualization techniques. We also present some empirical studies on real patient data from a large healthcare network over 200K patients. Finally, we envision the patient similarity framework can enable many important clinical applications such as comparative effectiveness research (CER), treatment recommendation, and physician comparison model.
Bio:
Jimeng Sun received the Ph.D. degree in Carnegie Mellon University in 2007, and is currently working at the IBM TJ Watson Research Center. His research interests include: medical data mining, large-scale data mining, web mining, visualization analysis. He won the ICDM2008 Best Paper Award, SDM2007 Best Paper Award, KDD2007 the best doctoral dissertation the second prize. He has published over 60 international journal and conference papers, and authorized the four patents. He also Serves as Chairman of a number of international Workshop, and more than 20 international conferences program members (including: SIGKDD, ICDM, SDM, PKDD and CIKM).