Table 3 Predictive accuracy of the combined CNN-MLP algorithm without and with bias correction.

From: Brain age prediction using combined deep convolutional neural network and multi-layer perceptron algorithms

Input

Without bias correction

With bias correction

MAE (years)

RMSE (years)

R2

MAE (years)

RMSE (years)

R2

Internal validation

(Test set, n = 301)

3.184

4.687

0.936

3.134

4.510

0.941

External validation

(CamCAN set, n = 645)

4.910

6.148

0.891

4.313

5.546

0.911

  1. CamCAN Cambridge Centre for Ageing and Neuroscience, MAE mean absolute error, RMSE root mean squared error, R2 coefficient of determination.
Back to article page