Traces of blood: Russia has created software for early cancer diagnosis
Sechenov University has registered software for the early diagnosis of cancer. Oncopro will be able to identify patients at risk by blood tests — just upload the results to the website. Izvestia found out how the development of Russian scientists works and for whom it is intended.
Important development
Sechenov University has registered Oncopro— a software for early cancer diagnosis. Izvestia was told about this at the university.
The Oncopro system can determine the probability of having six oncological diseases based on a blood test at the level of cancer markers — lung cancer, kidney cancer, breast cancer, ovarian or prostate cancer, colorectal cancer. A total of 12 indicators are involved in the calculation. The software has been tested at the clinical center for several years, which has made it possible to verify its accuracy and reliability.
— This is the result of several years of work by a large team of doctors, biologists, and mathematicians. We tried to make the product reliable, simple, and easy to learn," said Nikolai Agafonov, head of the development team and researcher at the Institute of Personalized Oncology at Sechenov University.
Marina Sekacheva, MD, Project manager, Director of the Institute of Personalized Oncology at Sechenov University, added that the system will become a convenient tool for determining the likelihood of cancer.
— Thus, the doctor will be able to make a timely decision on an in—depth examination of patients from risk groups, which will eventually help identify more cases of cancer at an early stage, - she is sure.
The Sechenov University Institute of Personalized Oncology worked on the product jointly with the V.P. Ivannikov Institute of System Programming of the Russian Academy of Sciences as part of the program for the creation and development of the NCMU "Digital Biodesign and Personalized Healthcare". According to the developers, in the future, the number of cancers, the risk of which is determined by Oncopro, can be expanded, work on the system continues.
How Oncopro works
As Marina Sekacheva, the project manager, explains to Izvestia, Oncopro was created on the basis of three machine learning algorithms. The developers created them by working with the blood counts of 600 volunteers, about 400 of whom had cancer. Now the system can receive the results of human tests, compare them with the results of cancer patients and assess the risk of developing the disease.
— A person takes a complex of blood tests, which includes cancer markers, indicators of inflammation, metabolism, and the coagulation system, which change with cancer. Then they are uploaded to the website, and the program outputs a risk calculator — high or uncertain. Patients with high blood pressure need additional examination, to which they will be referred by a doctor,— said the director of the Institute of Personalized Oncology at Sechenov University.
According to her, there is no such software in the country, but similar developments were conducted in China and South Korea. This indicates a global trend towards a risk—stratification model for cancer detection - when a patient is tested for not one, but several indicators at once — the so-called multitests.
— It is obvious that one cancer marker does not allow to make a diagnosis and even more so to understand the localization, where the tumor is located. The sensitivity and specificity of such programs, also using modern processing methods, that is, using machine learning, is a trend, and we are on trend," the expert noted.
Application prospects
According to Marina Sekacheva, the system was created last year, and this year the development team conducted technical and clinical trials to obtain a registration certificate. The software can now be used by medical institutions.
— The program is in the cloud, and, in all likelihood, we will provide colleagues with access to it, — said the interlocutor of Izvestia.
According to the Doctor of medical Sciences, in the future, the development can be used for testing and early diagnosis of various types of cancer. Screening procedures are complex, expensive, and not always effective manipulations, so modern scientists are striving to adhere to a new model that will allow them to identify risk groups and work with them specifically.
— We assume that it is easier to take a blood test than to perform a manipulation. And then the program will identify those who have the highest probability of cancer, so as not to "miss" the disease in the future. In general, it is designed for healthy people over the age of 40," added Marina Sekacheva.
At the same time, the Doctor of Medical Sciences emphasized, the program has limitations.: It does not work with patients with acute chronic inflammatory diseases, as it can produce false positive results.
Step into the future
According to oncologists interviewed by Izvestia, the development of Russian scientists can be a significant step in predictive medicine. The fact that it has already been registered as a medical device indicates its reliability and gradual readiness to be introduced into clinical practice.
— The main advantage of Oncopro is its preventive focus and early detection of oncopathologies. We all know that the detection of malignant neoplasms in the initial stages significantly increases the chances of successful treatment and reduces mortality. Unlike traditional methods, which are often used with already pronounced symptoms, this system in the future will be able to allow us to assess individual risk, which is especially important for patients without obvious clinical manifestations," said Anastasia Platonova, mammologist, chemotherapist and Oncologica Foundation expert.
In addition, the doctor added, early diagnosis will reduce the burden on the healthcare system, since the treatment of advanced forms of cancer requires significantly more resources. The introduction of such technologies into screening programs can become an effective tool for primary and secondary prevention, allowing doctors to promptly refer patients at risk for in-depth examination.
— Oncopro is currently focusing on the six most common types of cancer, and there are already proven cancer markers for them, such as PSA or CA-125. But I am sure that in the future the range of pathologies examined will only expand, and the use of this method in telemedicine will allow remote monitoring of risk groups," the oncologist believes.
In turn, Maxim Tersky, PhD, Chief physician of the Department of Clinical Laboratory Diagnostics and Endoscopy at JSC "Medicine" (Academician Roitberg Clinic), noted that the development could become a digital assistant for oncologists, which would save time on analyzing the obtained laboratory data. The emergence of such technologies in the field is closely monitored and the best of them are being implemented.
In general, AstraZeneca tells Izvestia that the widespread adoption of AI has become one of the key trends in innovative healthcare. Technology dramatically reduces the cost of conducting massive disease screening and prevention programs, while increasing the accuracy of clinical and laboratory diagnoses. For example, in 2021, with the support of the company, a project for the retrospective analysis of CT images using artificial intelligence was launched. It was held in Nizhny Novgorod and St. Petersburg. The AI analyzed the images originally taken to diagnose COVID-19 for the presence of tumors in the lungs. As a result, 113 cases of suspected lung cancer were identified in Nizhny Novgorod, and 158 such cases in St. Petersburg. Thus, retrospective screening using AI allowed to increase the potential level of diagnosis by 10%.
One way or another, experts conclude, in the future, Russian software may become a tool that can radically change the approach to screening and, most importantly, make it possible to detect the disease at the stage when treatment is most effective. For the patient, this means not only a chance for a cure, but also the preservation of the quality of life. Widespread implementation of the program may not happen soon, but the development is promising, so the probability of this is high.
Переведено сервисом «Яндекс Переводчик»
