Immersive Image Mining in Cardiology

Abstract

Buried within the human body, the heart prohibits direct inspection, so most knowledge about heart failure is obtained by autopsy (in hindsight). Live immersive inspection within the human heart requires advanced data acquisition, image mining and virtual reality techniques. Computational sciences are being exploited as means to investigate biomedical processes in cardiology. IntraVascular UltraSound (IVUS) has become a clinical tool in recent several years. In this immersive data acquisition procedure, voluminous separated slice images are taken by a camera, which is pulled back in the coronary artery. Image mining deals with the extraction of implicit knowledge, image data relationships, or other patterns not explicitly stored in the image databases (Hsu, Lee, & Zhang, 2002). Human medical data are among the most rewarding and difficult of all biological data to mine and analyze, which has the uniqueness of heterogeneity and are privacy- sensitive (Cios & Moore, 2002). The goals of immersive IVUS image mining are providing medical quantitative measurements, qualitative assessment, and cardiac knowledge discovery to serve clinical needs on diagnostics, therapies, and safety level, cost and risk effectiveness etc.