A Medical Scanning Device for Early Detection of Skin Cancer

Abstract

One fatal kind of cancer that can be successfully treated if detected early is Melanoma skin cancer. In this paper, an embedded diagnostic device to aid in the early identification of Melanoma is proposed. The reconfigurable computing advances and modern design approaches are used to construct an accurate and efficient medical device for identifying possible skin cancer. The aim is to develop and implement hardware and software modules that collaborate to diagnose skin cancer in images by analyzing them for possible tumors. This study employed the ResNet50 deep learning model, which produced a 78% accuracy rate. Additionally, experimentation with the VGG-16 model was conducted; however, it achieved a 75% accuracy rate and was thus disregarded in favor of ResNet50. The ResNet50 model was effectively installed on a Raspberry Pi3 Model B+, demonstrating its viability for real-world use. Testing and experimentation on the device's functionality reveal that it offers a potential solution for improving the skin cancer detection process' speed and accuracy. From the experimental outcomes, it is concluded that the proposed research has the potential to save lives and significantly advance the area of early skin cancer identification.