Skin detection is so deadly and when left unattended there is more possibility of death. About 9500 people in US alone is being diagnosed with skin cancer every day. Most of the people leave it unattended at early stages thinking it is just a normal skin burn or something. what if we bring a technology that can be used in any smart phones or computers and they can predict the severity easily. We have so much data of patients with which we can predict the cancer with images. With the use of deep learning we can do it.
Ours is a modern responsive webapp. It has 2 functions. The first function is you can take a image of your damaged skin and upload it to our server. The server will predict the possibility as Malign or benign and return it to the server. It uses a trained and deployed deep learning models with 75% accuracy. Once it returns the prediction, the user has the ability to book appointments with a doctor. The second function is we can directly request appointments from any doctor. It has a doctors page, where doctors can register for our service and can approve the appointment for further meetings with the patient for in-person tests
We used Flask as backend and keras is used for deep learning. SQLAlchemy is used for database creation. SQLite is the database used.
Model training is very hard within a day. Failed to deploy a accurate model. Time constraits
Building a proper web application in a day, Building something good for society which can be used by anyone as easy using social media.
Team work, learning about social issues
developing a smart phone based app. better implementation of the models