Inspiration

Skin cancer is a major public health problem, with over 5,000,000 newly diagnosed cases in the United States every year. Melanoma is the deadliest form of skin cancer, responsible for an overwhelming majority of skin cancer deaths. In 2015, the global incidence of melanoma was estimated to be over 350,000 cases, with almost 60,000 deaths. Although the mortality is significant, when detected early, melanoma survival exceeds 95%.

Millions of people across the globe suffer from skin cancer. Most of them are not able to detect it at an early stage, some due to ignorance, some due to financial situations. etc Our sole purpose to do exactly the same.

What it does

Detect skin cancer (melanoma) with a 75-90% accuracy.

How I built it

1] Got data for Skin Cancer from https://isic-archive.com/#images 2] InstalledTensorflow 3] Segregated Data on Basis of Benign and Malignant 4] Trained a CNN (convolutional neural network) for the segregated data 5] Created a python Cherry Pi Server as a mediator 6] Created a Web Interface using HTML/CSS/Javascript for a front end

Challenges I ran into

The cherry pi server on python could not establish a continuous connection to the tensorflow library. Had to write a different python script to establish the same and edit the label file from the tensorflow codebase.

Accomplishments that I'm proud of

I'm proud of not attempting to incorporate any of the company's APIs just to be eligible for their prizes because that would have caused unnecessary complexity and distracted me from achieving the main goal.

What I learned

Have patience. The code is not working because you didn't code it to work. Different levels of Skin Cancer (melanoma) Causes and how doctors might address it.

What's next for Mel-Detect

Raise money & publish on the internet for the world to use as a preliminary test for detecting skin cancer (Over the cloud, probably using AWS Web Services)

Built With

Share this project:

Updates