Scientists at South Ural State University (part of Project 5-100) have developed an application to effectively determine the likelihood of psoriasis using neural networks.
What is psoriasis and why is it difficult to diagnose?
Currently, psoriasis is one of the most common chronic recurrent skin diseases, affecting 1% to 5% of the population in different countries.
It is a chronic inflammatory skin disease. Psoriasis is a genetic disease (inherited), but it often manifests itself after severe stress or illness. Skin lesions are formations on the patient’s body: papules or plaques of pink-red color, the surface of which is flaky and covered with scales. The disease is accompanied by severe itching and pain.
It is difficult to correctly diagnose the form and type of psoriasis due to its similarity with many other diseases. Various analyzes and a comprehensive examination are required. Only a qualified and experienced dermatologist can make the correct diagnosis and prescribe treatment in order to achieve a long period of remission and avoid damage to internal organs.
How does the new technology work?
Scientists of the Department of System Programming of the Higher School of Electronics and Computer Science under the guidance of Candidate of Technical Sciences, Associate Professor Mikhail Sukhov worked on creating a program that will accurately diagnose psoriasis.
To train neural networks, the method of teaching with a teacher was chosen, where the learning algorithm itself acts in its role. Learning is based on a sample, which is a dataset of two types of images: with and without psoriasis. A preprocessed image is fed to the input of the neural network. The output is the classification results and the accuracy of this result.
“Determination of psoriasis from photography using a neural network approach will reduce the likelihood of dermatologists making mistakes in determining the type of disease,” says Mikhail Sukhov. – In modern science and medicine, the topic of using neural networks to determine the presence of a disease is very popular. Such systems have several advantages over the expert physician. They are objective, stable, and provide an optimal solution based on a huge knowledge base that they retain forever. At the same time, this approach completely excludes the human factor, because the knowledge of even a competent specialist can be forgotten or confused. ”
As a result, the neural network allows you to analyze the skin from photographs and, if there are signs of a disease, to report it.
What are the prospects?
Further research in this subject area suggests improving and replenishing the database, which will increase the accuracy of image classification, reduce the likelihood of errors in the presence of images with other outwardly similar skin diseases.
It will also be possible to teach the neural network to recognize the types of psoriasis. The more baseline data with different types of psoriasis, the higher the likelihood of a possible diagnosis of the disease. But to solve such a problem, closer cooperation with specialists in this field is required.