Detection of Skin Conditions from a Photo Using AI
AI-based online service and mobile application that recognize 7 types of skin rashes from photos: melanomas, basaliomas, melanocytic nevi, actinic and seborrheic keratoses, vascular rashes and dermatofibromas
How does it work?
1. A patient or doctor uploads a photo via an online service or mobile application.
2. The algorithm analyzes them and compares them with the sample base for diagnostics.
3. The system generates a suggested diagnosis that must be verified by a doctor.
The service helps doctors and patients to
Reduce the risk of medical errors
Accelerate analysis of symptoms for initial diagnosis
Get affordable self-diagnosis for faster detection of skin conditions
Reduce the burden on doctors through primary self-diagnosis
Detailed operating principle
Skin AI is based on the algorithms of the convolutional neural network EﬃcientNet, pre-trained on the ImageNet dataset, which is responsible for the analysis and classification of images. Data for training provided by Cornell University (New York, USA).