Skip to main content
. Author manuscript; available in PMC: 2019 Apr 8.
Published in final edited form as: J Xray Sci Technol. 2019;27(1):17–35. doi: 10.3233/XST-180426

Fig. 4.

Fig. 4.

Our Basic CNN model Design, which contains two convolution layers with kernel size 5 × 5 and 3 × 3 respectively, each with LeakyReLU as activation function and followed by a Max-pooling layer. The output layer has two nodes representing the two classes: nodule and non-nodule.