Fig. 5From: Shape-based disease grading via functional maps and graph convolutional networks with application to Alzheimer’s diseaseThe proposed GCN approach achieves the highest average classification accuracy of \(0.787 \pm 0.055\) followed by CNN (\(0.77 \pm 0.06\)), MCNN (\(0.767 \pm 0.038\)) and MLP (\(0.749 \pm 0.074\)) [16, 17]Back to article page
