Comparison and inference of white matter hyper intensities on MRI
It is still unclear what causes white matter lesions (WML) of the brain. However, it has been known that WML are consistently associated with aging and hypertension. Moreover, several chronic disorders such as Alzheimer's disease and depression have been found to be associated with WML. Therefore, understanding complex patterns of WML may contribute to answering important questions on the causation and consequences of WMH, which leads to early prediction of disease and preventive treatment. This project attempted to study the similarity of white matter lesions of the brain on the MRI. We have recently proposed an original computational neuroscience framework for such a conceptual study with a standpoint that the prior knowledge about WML accumulated and utilized to enable a reliable inference of the rating of a WML observation. This computational approach for rating inference of WML, which appears to be the first study, can be utilized as a computerized rating-assisting tool and can be very economical for diagnostic evaluation of brain tissue lesions.The preliminary results of this project appear to be promising for clinical application and worth further investigation both technically and clinically.