Multimodal hybrid convolutional neural network based brain tumor grade classification

BMC Bioinformatics

Table 4 The detailed architecture of Visual Geometry Group-16 and Visual Geometry Group-19

No. of layer	Size of the output	Visual geometry group-16	Visual geometry group-19
Convolution block 1	Size of 224 × 224	conv 2D × 2	conv 2D × 2
Convolution block 1	Size of 112 × 112	2D max_pooling layer	2D max_pooling layer
Convolution block 2	Size of 112 × 112	conv 2D × 2	conv 2D × 2
Convolution block 2	Size of 56 × 56	2D max_pooling layer	2D max_pooling layer
Convolution block 3	Size of 56 × 56	conv 2D × 3	conv 2D × 4
Convolution block 3	Size of 28 × 28	2D max_pooling layer	2D max_pooling layer
Convolution block 4	Size of 28 × 28	conv 2D × 3	conv 2D × 4
Convolution block 4	Size of 14 × 14	2D max_pooling layer	2D max_pooling layer
Convolution block 5	Size of 14 × 14	conv 2D × 3	conv 2D × 4
Convolution block 5	Size of 7 × 7	2D max_pooling layer	2D max_pooling layer
Classification layer	4096	(softmax layer\|fully connected layer) × 3	(softmax layer\|fully connected layer) × 3

ISSN: 1471-2105