Snehashis Roy1, Aaron Carass1,
Navid Shiee1, Dzung L. Pham2, Susan Resnick3,
Jerry L. Prince1,2
1Electrical and Computer Engg, Johns
Hopkins University, Baltimore, MD, United States; 2Department of
Radiology and Radiological Science, Johns Hopkins University, Baltimore, MD,
United States; 3Laboratory of Personality and Cognition, National
Institute on Aging, Baltimore, MD, United States
Progression
of white matter lesions are important for early detection and monitoring of
diseases like Alzheimer's or Multiple Sclerosis. FLAIR images provide
superior contrast for lesions compared to traditional T1 or T2 weighted
images. But they are often not acquired for time and cost constraints. We
developed an atlas based method to synthesize FLAIR images from T1 and T2
acquisitions. We use this method to quantify the progression of lesions on a
pool of 20 subjects. Synthesizing FLAIRs can be seen as a potential way to
reduce unnecessary data acquisition.