A Survey on Human Cancer Categorization Based on Deep Learning

Abstract

In recent years, we have witnessed the fast growth of deep learning, which involves deep neural networks, and the development of the computing capability of computer devices following the advance of graphics processing units (GPUs). Deep learning can prototypically and successfully categorize histopathological images, which involves imaging classification. Various research teams apply deep learning to medical diagnoses, especially cancer diseases. Convolutional neural networks (CNNs) detect the conventional visual features of disease diagnoses, e.g., lung, skin, brain, prostate, and breast cancer. A CNN has a procedure for perfectly investigating medicinal science images. This study assesses the main deep learning concepts relevant to medicinal image investigation and surveys several charities in the field. In addition, it covers the main categories of imaging procedures in medication. The survey comprises the usage of deep learning for object detection, classification, and human cancer categorization. In addition, the most popular cancer types have also been introduced. This article discusses the Vision-Based Deep Learning System among the dissimilar sorts of data mining techniques and networks. It then introduces the most extensively used DL network category, which is convolutional neural networks (CNNs) and investigates how CNN architectures have evolved. Starting with Alex Net and progressing with the Google and VGG networks, finally, a discussion of the revealed challenges and trends for upcoming research is held.

Keywords: cancer types; convolutional neural network; deep learning; human cancer; medical imaging.

Figure 1

Key component of building a (B) deep belief network is a (A) restricted Boltzmann machine (A) (Ghosh et al., 2021).

Figure 2

(B) Stacked autoencoders are built by (A) stacked autoencoders in sequence, best viewed in color (Xiang et al., 2016).

Figure 3

Different feature maps of a convolutional neural network [CNN, simpler from the left and more semantic (comprehensive) going (deeper) to the right] (Sarvamangala and Kulkarni, 2021).

Figure 4

Key element of CNN architecture is the convolutional layer that involves convolution, activation, and pooling (Sarvamangala and Kulkarni, 2021).

Figure 5

Average percentage of annual deaths due to various cancer types, best viewed in color (Munir et al., 2019).

Figure 6

Qualitative difference between the 2D and 3D (tomosynthesis) mammography for breast cancer diagnosis (Zhou et al., 2019).

Figure 7

Various DL approaches' recent performances in terms of their diagnosing accuracy for (A) lung cancer, (B) breast cancer, (C) prostate cancer, (D) brain cancer, and (E) skin cancer (Munir et al., 2019).

Figure 8

Typical flow graphs for the pre-processing tasks of well-known medical images: MRI, computed tomography (CT), mammogram, and transrectal ultrasound (TRUS) are illustrated from top to bottom, best viewed in color (Elazab et al., 2020).