Fig. 1From: 3D convolutional neural networks uncover modality-specific brain-imaging predictors for Alzheimer’s disease sub-scores3D CNN model architecture. The CNN model is composed of 10 convolutional layers grouped: Conv 1 (2 layers), Conv 2 (2 layers), Conv 3 (3 layers) and Conv 4 (3 layers). This model processes brain-imaging data and can predict both AD diagnosis (yellow box) and ADAS–Cog13 sub-scores (grey box)Back to article page