Russia has developed an IT solution to identify people at high risk of cancer
The laboratory of molecular genetic research at Berezin Sergei Medical Institute is completing testing of the Russian development of Onqueta. This is a digital solution for identifying people with high cancer risk for subsequent genetic testing.IT-product of domestic developers Onqueta has been successfully tested at the laboratory of molecular genetic research of Berezina Sergei Medical Institute. With the help of a "question and answer" system, without the participation of a doctor, a special algorithm evaluates the probability of hereditary cancer syndrome and gives recommendations for further genetic testing, said "MV" the scientific consultant of the project, oncologist Marat Gordiev.
Onqueta is a domestic analogue of the most advanced, according to experts, screening system using AI CancerIQ. It is a questionnaire that the patient can take at home, in a quiet environment, and then his answers are processed by artificial intelligence. The algorithm then determines the likelihood of a hereditary cancer syndrome and makes recommendations about the need for genetic testing.
"Not every cancer is hereditary. The incidence of carriers of hereditary syndromes is only 1.5-2%. But to know for sure, you have to have genetic testing. The MHI does not pay for it, but it is not cheap. This is where inexpensive digital solutions come to the rescue," Gordiev explained.
Previously, the doctor needs to establish who in the patient's family had cancer, at what age, when they passed away. To do this during the appointment is problematic. At home, a person can spend more time on the questionnaire without taking it away from the doctor.
The service is supposed to work in the public domain. At the same time, the algorithm will be able to ask follow-up questions if it lacks information for analysis. If the AI detects a high risk of genetic predisposition to cancer, the patient may be recommended to undergo testing to verify the prognosis.
During Onqueta testing, 1,551 cases from the database "Hereditary cancer syndromes in the Russian Federation" were divided into two groups: with pathogenic mutations (731 cases) and without them (820 cases). In the first group, 690 cases (94%) could be detected by the appendix; in the second group, 800 of 820 cases were classified by the algorithm as having a low probability of detecting hereditary cancer syndromes. Thus, the app had a sensitivity of 94% and specificity of 97.5%, Gordiev stated.
Now the app's front-end is being fine-tuned. The MVP is expected to be ready soon.