A Medical Support Application for Public based on Convolutional neural network to Detect Skin Cancer

Health problem, as the most currently wide social consensus, has attracted great attention in the medical domain and academic community. Especially most current females, they concern about their skin problems even more than average people. On the one hand, however, people may unconsciously ignore their skin problems since the peace of their work and life is getting faster and faster today. On the other hand, it is still hard for people to take time out to do a series of professional tests to discover the potential skin problems by going to hospitals even though they can notice the uncomfortable feeling in touching or the unnormal color surface of the skin. To alleviate the worry of people's skin problems effectively and warn them of the problems in time, we designed a medical support application for the public and community based on Convolutional Neural Network to detect common skin cancers in general. In this application, we trained a convolutional neural network that can effectively extract the medical knowledge that it can analyze whether the inputs, as known as skin images or non-skin images at first. The technology should analyze which kinds of skin problems from the skin images father. In addition, this application provides two extra supported functions, including the chat room and the personal center. A user can ask questions via the chat room that there is an AI assistant who would respond to the questions properly. It is also available for the user to contact a real doctor by typing the specific content. What's more, the personal center allows logged-in users to browse the history of tests and take feedback for the certain test based on the professional result of the skin problem.