Project Hiraya

• 

  Patient inputs their profile

• 

  Answers questions about symptoms

• 

  Inputs duration of symptoms

• 

• 

  Patient selects way of adding photo

• 

  Patient adds Photo of their skin

• 

  Model Training using test data (600+ sample images)

• 

  Model training using Training data (2000+ sample images)

• 

  Sample images of skin cancer

• 

  Accuracy results of the model

Inspiration

• Thought about how cool it could be to diagnose the skin problems without taking any test from fancy medical lab The amount of radiation reaching earth along with the depletion of the ozone layer has been increasing global temperatures. With this, the incidence and rate of development of skin cancer have also been rising. Currently, to detect skin cancer, a person needs to go to a licensed dermatologist to get a biopsy. These visits can be expensive and out of reach for many individuals. We saw the need to make skin cancer analysis faster, more affordable, and accessible. That is why we developed the Skin Cam ## What it does Diagnose your skin with a single picture, and send your results to your dermatologist!

Front end

IONIC: Used to build the mobile application to gather data from the user

Back end

We used image samples of benign and malignant skin cancer. We used Scikit Learn as the library for our Machine learning algorithm.

Try It here: ( https://github.com/HirayaIBMZ/hiraya )

Challenges we ran into

The Z platform does not support tensor flow

Built With

• angular.js
• ionic
• machine-learning
• python