Bioinformatics and Machine Learning in Skin Cancer Risk Assessment and Prognosis: A Review

Abstract

Skin cancer is one of the leading dangerous varieties of cancer. Carcinoma Skin cancer is caused by unrepaired deoxyribonucleic acid (DNA) in skin cells that produce mutations or genetic defects on the skin. Carcinoma has a tendency to moderately extend to other parts of body, therefore it is easy to cure during early stages, and hence it is required to detect as early as possible. Every year, doctors diagnose carcinoma in around three million or more patients across the world. Nowadays, it is one of the most widely recognized forms of cancers for human health. Hence, we need an early diagnosis to prevail any crucial condition of the infected patients. There a lot of factors such as the rate of increase of cases, increased death rate, and more expensive and painful medical treatment. Having considered the seriousness of those problems, researchers have developed numerous early finding methods for skin cancer. Despite clinical staging guidelines, the prognosis skin cancer (metastatic melanoma) is variable and difficult to predict. Machine Learning and Bioinformatics take inputs from clinical, histopathology and genetic to analyze to predict risk with high accuracy of melanoma patients. This literature review aims to provide key genetics science drivers of malignant melanoma and up to date applications of machine learning models and bioinformatics with the risk discovery of carcinoma patients. A robustly valid risk stratification tool will probably guide the medical practitioner management of malignant melanoma patients and ultimately improve patient outcomes. Review findings are presented in tables for better understanding.