Skip to main content
. Author manuscript; available in PMC: 2020 Oct 1.
Published in final edited form as: Magn Reson Imaging. 2019 Jun 24;62:70–77. doi: 10.1016/j.mri.2019.06.018

Fig 1.

Fig 1.

Pipeline for age prediction. Two sets of features are used: intensity-derived features (red) derived from a convolutional neural network of increasing filter size (red boxes), and structural features (blue) using multiatlas segmentation (bottom). These features are concatenated and used as inputs to directly predict age. BN: Batch Normalization; ReLU: Rectified Linear Unit Activation; Max Pool: Max Pooling Layer.