AI learned to diagnose skin diseases

AI has learned to diagnose skin diseases. The system will help medical professionals more accurately diagnose 134 different skin diseases. The work of researchers from Seoul National University is published in the journal Journal of Investigative Dermatology.

Skin diseases are widespread, but even dermatologists are not always able to distinguish benign from malignant diseases at once.

In order to solve this problem and help doctors more accurately diagnose, the researchers created an AI-based on the convolutional neural network that can accurately detect malignant neoplasms, propose treatment options and classify skin diseases.

The neural network was trained on 220 thousand images of diseases on the skin of people of different nationalities. The resulting AI was able to diagnose 134 skin diseases and offer primary treatment options.

“Our work shows that AI can serve as an assistant that can expand the capabilities of medical professionals in the field of diagnostic dermatology. Instead of replacing people with AI, we expect AI to support people because augmented intelligence will make diagnoses faster and more accurate”.

Jung-Im Na, lead author of the study

Google News button