MOSCOW, March 20A method for identifying skin tumors based on non-invasive studies followed by mathematical analysis using artificial intelligence was developed by scientists from SSMU. According to them, the method allows you to accurately determine the stage of tumor development and choose the right treatment tactics. The results were published in the journal Diagnostics.
Basal cell carcinoma of the skin (basal cell carcinoma) is one of the most common types of skin tumors worldwide. Correct classification of skin lesions is a key screening step that requires high accuracy and interpretability.
The risk of developing basal cell carcinoma (BCC) increases in the presence of factors such as genetic predisposition, low melanin content in the skin, prolonged exposure to ultraviolet radiation on exposed areas of the body, etc. According to the European Consensus Interdisciplinary Guidelines for the Diagnosis and Treatment of Basal Cell Carcinoma, within the next 10 years, the incidence of BCC will increase by 30% among men and by 25% among women.
Scientists from Saratov State Medical University named after. V.I. Razumovsky (SSMU) conducted a study aimed at improving the treatment of such patients. They developed a method for identifying skin tumors based on non-invasive studies followed by mathematical analysis using artificial intelligence.
“This is important for the preoperative classification of the tumor and determining the degree of its invasion, which makes it possible to select the optimal treatment tactics — surgical treatment, laser coagulation, photodynamic therapy, close-focus X-ray therapy,” noted Sergei Kapralov, head of the department of faculty surgery and oncology of SSMU, adding that Based on the data obtained, it is possible to clarify the scope of surgical intervention.
In addition, he estimates that the results of the study will shorten the diagnostic stage in patients with skin tumors and reduce the risk of misdiagnosis.
As the university said, similar studies have already been conducted around the world, but in each of them only one of the methods for diagnosing skin tumors was used. Saratov scientists, according to them, analyzed a whole combination of non-invasive methods — dermatoscopy, ultrasound scanning (US), optical coherence tomography (OCT) and diffuse reflectance spectroscopy of the tumor and healthy skin surrounding the tumor. Based on the data obtained, they developed a machine learning algorithm that allows them to identify skin tumors and determine further treatment tactics for patients.
““Dermatoscopy clarifies the nature of the pathological process, studies the surface of the tumor and evaluates changes in the surrounding tissue under tenfold magnification,” explained Kapralov.
He added that ultrasound and OCT of the skin assess the degree of tumor invasion, the nature of pathological changes in tissue in the tumor area and surrounding tissues.
Currently, scientists are faced with the task of developing algorithm for choosing a treatment method for patients with skin tumors in order to prevent relapse and implement the obtained data in practical healthcare.
The research was carried out as part of joint scientific work with the Department of Optics and Biophotonics of SSU named after. N.G. Chernyshevsky with the support of the Ministry of Education and Science of the Russian Federation and a grant from the Russian Foundation for Basic Research.