In the US, a neural network has been created to detect caries in dental images.

It will simplify the work of dentists and will become the basis of mobile applications for initial self-diagnosis. She was trained on photographs of the upper and lower jaws of several hundred patients. The algorithm identified the destructive process with an accuracy of 79.5%. He managed to find more than 83% of the affected teeth and correctly determine the boundaries of caries in almost 80% of the enamel areas.