Contemporary Trends in the Early Detection and Diagnosis of Human Cancers Using Deep Learning Techniques
Nirmala Vasan Balasenthilkumaran, Sumit Kumar Jindal · 2023
According to the World Health Organization (WHO), cancers are one of the leading causes of death and have accounted for nearly 10 million deaths worldwide in 2020. This number is only projected to increase in the future. Early detection and diagnosis of cancers would help curb the mortality rate by increasing chances of survival and allowing early interventions and the use of inexpensive treatment methods. Deep learning (DL) has been successfully used in the medical industry in areas such as oncology, radiology, cardiology, neurology, urology, and osteology for medical image analysis and object recognition. Recently, convolutional neural networks (CNNs), autoencoders, and artificial neural networks (ANNs) have begun to be utilized to develop highly accurate algorithms and have been coupled with computer-aided diagnostic techniques to aid clinicians in cancer diagnostics. This chapter analyzes the use of DL architectures for early detection and diagnosis of human cancer. The recent progress involving the use of different DL techniques for the early diagnosis of various cancers is discussed, the current limitations and challenges for the implementation of DL techniques are assessed, and solutions to some of the common challenges faced by researchers are presented. Finally, directions for future research are outlined.