European Journal of Molecular & Clinical Medicine

Abstract

The prevalence of malignant melanoma is increasing worldwide. This cancer is one of the leading causes of death in young people at any age. This cancer can be identified very early on because it's curable because it is noticeable on the face. New technologies have been a possible possibility to completely accurately diagnose early melanoma. First, a significantly better scientific detection capacity than melanoma in the clinic to be observed at the very earliest point with the advent of dermoscopy.The global implementation of this method has allowed vast collections of dermoscopic photographs of histopathologically confirmed melanomas and benign lesions to accumulate. In the fields of image recognition and machine learning, the advancement of advanced technology has allowed us to differentiate malignant melanoma from the several innocuous fake pictures without any biopsy. Such modern techniques would not only enable melanoma to be diagnosed sooner but should also reduce the large number of unnecessary and costly biopsy procedures.Although some of the latest systems in preliminary trials demonstrated a potential for these innovations, more technological improvement must be anticipated in precision and reproductiveness. We offer an outline on melanoma computerized detection in dermoscopic images in this article. First, the various types of lesion segmentation are discussed. We then offer a brief description of the segmentation of clinical characteristics. Finally, we are addressing the classification process in which machine learning algorithms are used to predict melanoma attributes produced by segments.