Introducing AI technologies for dementia diagnosis Lead Investigator: Akihiko Shiino Institution : Shiga University of Medical Science E-Mail : shiino@belle.shiga-med.ac.jp Proposal ID : 1366 Proposal Description: We are developing AI trained by ADNI database, and validated by AIBL and JADNI (MRI). Therefore we would like to check the performance on the NACC data. We are preparing to publish our data, and we found a recent elegant report from Dr. Kolachalama (http://dx.doi.org/10.1101/832519) which used NACC data. NACC database are built on different criteria from ADNI, and it is really practical to compare the AI performance to the other. Our final purpose is to introduce the AI to practical clinical use to support early diagnosis from MRI. Considering the rapid increase of dementia patients and being a shortage of specialists at present, we believe that the AI can reduce the reliance on personal abilities of clinicians by offering a likelihood of the disease in terms of brain atrophy.