Figure 1. The overall workflow of our computational framework.

The main procedure included data acquisition, pre-processing and classification. A traditional machine learning algorithm with feature selection and two deep 3D convolutional neural networks for Alzheimers disease classification have been investigated in this study.