Certain investigations on the classification and prediction of melanoma using image segmentation based on deep learning techniques

Abstract

Melanoma is a serious type of skin cancer that begins in melanocytes, newlinethe cells that produce skin pigment. The detection of melanoma in the earlier newlinestage is very difficult, so it leads to very less survival rate of the patients. The newlinebenign skin lesions and the melanoma have same visual similarities, which newlineleads Melanoma is a life-threatening form of skin cancer that originates in the newlinemelanocytes, the cells responsible for producing pigment in the skin. The newlinedetection of melanoma in the earlier stage is very difficult so it leads to very newlineless survival rate of the patients. The benign skin lesions and the melanoma newlinehave same visual similarities, which leads to the physician very challenging in newlinediagnosis of melanoma. To enhance the diagnostic rate of the melanoma, newlineArtificial intelligence and Deep learning techniques were utilized to reduce the newlinesubjectivity of manual assessments. newlineIn particular, Deep learning technique such as Convolution Neural newlineNetworks (CNNs) plays an important role in various medical imaging task such newlineas skin lesions classification. CNN will automatically learn about the newlinehierarchical features of the images for detecting the complex patterns which newlinemay not be easily identified by the human observers. Deep learning model can newlineidentify the differences between malignant and benign lesions and also newlineassisting the physician in early detection of melanoma. newline