Date of Award
Doctor of Philosophy (PhD)
Srimani , Pradip K.
Luo , Feng
Pargas , Roy P.
The aging population has increased the importance of identifying and understanding mild cognitive impairment (MCI), particularly given that 6 - 15 % of MCI cases convert to Alzheimer's disease (AD) each year. The early identification of MCI has the potential for timely therapeutic interventions that would limit the advancement of
MCI to AD. However, it is difficult to identify MCI-related pathology based on visual inspection because these changes in brain morphology are subtle and spatially distributed. Therefore, reliable and automated methods to identify subtle changes in morphological
characteristics of MCI would aid in the identification and understanding of MCI. Meanwhile, usability becomes a major limitation in the development of clinically applicable classifiers. Furthermore, subject privacy is an additional issue in the usage of
human brain images.
To address the critical need, a complete computer aided diagnosis (CAD) system for automated detection of MCI from heterogeneous brain images is developed. This system provides functions for image processing, classification of MCI subjects from control, visualization of affected regions of interest (ROIs), data sharing among different
research sites, and knowledge sharing through image annotation.
Li, Lin, "A COMPUTATIONAL PIPELINE FOR MCI DETECTION FROM HETEROGENEOUS BRAIN IMAGES" (2012). All Dissertations. 882.