Scientists from the North Caucasus Federal University (NCFU) introduce a new development.
Researchers at the North Caucasus Federal University (NCFU) have created a neural network system designed to assist in diagnosing skin cancer. They propose using this developed system as an auxiliary diagnostic method. This will not only improve the accuracy of cancer detection but also reduce the influence of human factors on the final decision.
The neural network system developed at NCFU is sensitive to unbalanced dermatological data, thereby potentially reducing the number of false-negative forecasts.
According to the scientists, the recognition accuracy of the neural network system across ten diagnostic categories reached 85.2%. On average, this provides a 15% more precise outcome compared to the visual diagnosis performed by practicing physicians.