Page 258 - Handbook of Deep Learning in Biomedical Engineering Techniques and Applications

Chapter 9 Applications of deep learning in biomedical engineering 249

The two prominent properties of deep neural model are
(1) nonlinear processing units, and
(2) feature representations using supervised or unsupervised
learning [6].
ANNs and DL are the leading tools in several domains such as
image analysis and fault diagnosis.
The applications of the DL in the biomedical fields cover all
the medical levels, ranging from the omics, such as the protein
and gene expression, to the public medical health management,
such as predicting demographic information or infectious disease
epidemics [6].

4. Most popular deep neural networks
architectures used in biomedical
applications
• Convolutional neural network (CNN)
• Generative adversarial network (GAN)
• Recurrent neural network (RNN)
• Deep belief network (DBN)

5. Convolutional neural network
CNN is a particular type of feed-forward neural network in AI.
It is widely used for image recognition [7]. CNN represents the
input data in the form of multidimensional arrays [2]. It works
well for a large number of labeled data. CNN extract the each
and every portion of input image, which is known as receptive
field. It assigns weights for each neuron based on the significant
role of the receptive field. So that it can discriminate the impor-
tance of neurons from one another [8]. The architecture of CNN
is shown in Fig. 9.4. The architecture of CNN consists of three
types of layer: (1) convolution, (2) pooling, and (3) fully connected.

6. Convolution layer
The purpose of the convolutional layer is to obtain significant
features from the input. The convolutional layer convolved the
input image with set of kernels or filters and learnable parame-
ters. The layer generates the feature map by involving the
following steps.