Skin Cancer Lesions Classification Using Probabilistic Neural Network

Abstract

Abstract One of the most frequent types of cancer is skin cancer. Skin cancer is one of a deadly disease caused due to the abnormal skin cell proliferation in the epidermis resulting in the formation of mass called as Tumor. The formed tumor can be cancerous known as Malignant, that is it can raise and can open out to the rest of the body and the other is non-cancerous tumor known as benign which can raise but it does not open out to the rest of the parts. Skin cancer is mainly caused because of exposure of skin to the harmful UV(Ultraviolet) radiation. If skin cancer is detected in early stage then there are high chances for successful skin cancer treatment. In this paper we are describing a model which uses deep learning technique and some Image processing techniques for distinguishing of skin cancer as benign tumor or malignant tumor. The proposed model gives results in less time and eliminates the cost compared to the formal strategy that is Biopsy which is used for skin cancer recognition. The proposed model uses image processing techniques like Pre-processing, different Segmentation techniques, Feature Extraction and deep learning technique PNN (Probabilistic Neural Network) is used for classification of the skin image.