Fig. 4: Association between the shared disease effect and multiscale features using machine learning. | Communications Biology

Fig. 4: Association between the shared disease effect and multiscale features using machine learning.

From: Multiscale neural gradients reflect transdiagnostic effects of major psychiatric conditions on cortical morphology

Fig. 4

a Probability of the selected features across five-fold nested cross-validations and 100 repetitions for predicting the shared disease effect. The frequently selected features are reported with asterisks. b Linear correlation between actual and predicted values of the effects is shown on a scatter plot. The black line indicates mean correlation and the gray lines represent the 95% confidence interval for 100 iterations with different training/test datasets. SD standard deviation, FDOPA 18 F fluorodopa, DAT dopamine transporter, NAT noradrenaline transporter, SERT serotonin transporter, MAE mean absolute error.

Back to article page