An interpretable deep learning framework for Alzheimer's disease diagnosi

 
   Lead Investigator:    Chao Liu
   Institution      :    Peking University
   E-Mail           :    chaoliu_pku@pku.edu.cn
   Proposal ID      :    1512



   Proposal Description:
   Current methods integrate patient history, neuropsychological testing, and MR to identify likely cases, yet effective practices remain variably applied and lacking in sensitivity and specificity, we plan to propose an algorithm based on deep learning method which can classify different degrees of Alzheimer's disease (AD, MCI, NC) based on MRI imaging data. Majority of diagnosing methods based on deep learning focused on classifying on AD or NC, and usually overlook the MCI, but the diagnosis of MCI  is more necessary for the elderly since it can delay the further development of the disease if we can find it out early and pass the corresponding early treatment. We currently need to gather more patient data of MRI which can help us to train a robust diagnostic model.