843-500-5316 How it works About Contribute

Classify a skin mark with one picture

DermaAI is a non-profit project that leveraging cutting-edge Machine Learning technologies amis to detect skin cancer conditions in a single picture shot for example with a common phone.

Frame the mark

To get an accurate result, the photo taken should only include the mark, preferably in the center.

If the photo includes other marks such as moles or lentigo, the algorithm does not know which one it has to analyize!
Be sure to take the photo in a bright place, preferably under natural light, although also white light is fine. Stronly coloured lights might affect the color of the mark thus hindering the classification process.


Focus and zoom

The bigger and clearer the image of the mark, the easier it is for the algorithm to classify it.

Get the camera close to the mark as much as possible and then adjust the focus on the mark. If the image is blurry, it might be because the camera is too close to the skin and can not focus properly. Thus it might be better to get slighly less close to the skin and instead use the camera zoom.
The quality of the image is important, take the image with a proper phone/camera and with a firm hand.

Try out now the Beta version!

Image formats currently supported: jpg and png

Return to home