Identification of Alzheimer’s disease using a convolutional neural network model based on T1-weighted magnetic resonance imaging

Datasets

We used two datasets in this study: one from the ADNI and the other from the SNUBH. From ADNI, we included participants in both ADNI1 and ADNI2 who had 3.0 T T1-weighted images and were diagnosed as CN or mild AD (CDR of 0.5 or 1). For up-to-date information about the ADNI, see https://www.adni-info.org. From the SNUBH, we included AD patients and CN controls with T1-weighted images whose age, sex, and CDR were matched to the patients from the ADNI. However, we were unable to further match for education and cognitive level because participants from the ADNI were more educated and performed better on the Mini Mental State Examination (MMSE) than those from the SNUBH. In the case where a participant has multiple MRI scans from different timepoints, we selected only one MRI scan based on the participant’s age and diagnosis at the time of assessment. We selected the scan whose…

Read more…