Skin cancer is one of the most common types of skin cancer, representing at least 40% of all cancer cases. The survival rate of Skin Cancer patients is significantly increased when it’s detected earlier by dermoscopic images. Our problem is, in skin cancer detection doctors might face some problems such as being confused by skin cancer and other skin conditions symptoms and other multiple reasons. Causing a delay in the patient’s diagnosis and to other patients as well, which might lead to worst conditions or even death. That makes the current process less efficient and more expensive . After considering these problems we proposed a solution which is developing a model that will help Doctors and patients to detect skin cancer in early stages . So, when there is a way to diagnose skin cancer in fast and accurate process, we can avoid the circumstances of skin cancer, increase the survival rate and it will help the patients in saving the medical cost and treatment.
Our vision is to provide a fast, precise, and capable model that will accurately recognize the lesion areas and classify them to Malignant or Benign lesions. In this project we use the Deep Conventional Neural Network algorithm written in python programming language to help with the segmentation and classification. DCNN are the most used type to identify patterns in videos and images. So, we believe that DCNN will help us achieving our goal.