Estimating Gender and Age from Brain Structural MRI of Children and Adolescents: A 3D Convolutional Neural Network Multitask Learning Model

<div>Scatter plots between predicted and target ages. (a) Ages prediction on the test set from the best-performing model of ADHD-200 10-fold cross-validation. (b) The same best-performing model, which was trained with ADHD-200 data, evaluated across the full ABIDE-II data set. Note. <i>r</i>: Pearson’s correlation between predicted and target ages, MAE: the mean absolute error of the predictions, and <i>n</i>: the evaluated sample size.</div>

Computational Intelligence and Neuroscience

fig3

Figure 3

Figure 3: Estimating Gender and Age from Brain Structural MRI of Children and Adolescents: A 3D Convolutional Neural Network Multitask Learning Model 

Figure 3 | Estimating Gender and Age from Brain Structural MRI of Children and Adolescents: A 3D Convolutional Neural Network Multitask Learning Model 