Developed a platform for training medical algorithms
As it became known to Kommersant, the Ministry of Health, together with Rostec, created the first version of the federal platform for artificial intelligence (AI) for healthcare. It should give IT developers access to the impersonal medical data of Russians from electronic medical records. The goal of the project is to create AI services in the healthcare sector, the use of which can increase the speed of making diagnoses, reduce the number of medical errors, and even allow for ultra-early diagnostics of patients' predisposition to pathological changes.
Participants of the medical IT services market told Kommersant about the creation of a federal platform of artificial intelligence for the healthcare sector, and the Ministry of Health confirmed this information. The ministry said that the main goal of the project, which Rostec is currently working on, is to combine anonymized and labeled medical data into verified datasets (datasets), as well as provide Russian IT companies with a platform for developing and testing AI services in the healthcare sector. Companies need access to structured data to develop algorithms that can later become the basis of medical decision support systems. The Ministry of Health and the state corporation Kommersant were told that they are not yet ready to disclose the cost of creating such a platform.
The source of data for it will be, first of all, the subsystem of the Unified State Information System in the field of healthcare “Federal Integrated Electronic Medical Record”. Since the end of 2020, its functions include storing sets of anonymized medical data for machine learning algorithms and labeling them.
In addition, it is planned to use medical information accumulated in national medical research centers subordinate to the Ministry of Health as a data source. According to Rostec, now 18 out of 33 such institutions are involved in projects using AI or conduct their own research in this area. Including – the A.V. Vishnevsky National Medical Research Center of Surgery, the N.N.Blokhin National Medical Research Center of Oncology, the N.N. Blokhin National Medical Research Center of Endocrinology, the V.A.Almazov National Medical Research Center and the N.N.Burdenko National Medical Research Center of Neurosurgery.
The research of Rostec showed that now from the data collected by the National Medical Research Center, it is possible to form 14 archives for radiation diagnostics and 10 archives for pathomorphology. “On their basis, accordingly, datasets will be created for the interpretation of the data of radiation and histological studies, which will be placed on the platform first. These areas were chosen taking into account several criteria, first of all we looked at what would be most in demand for doctors, ”says a Kommersant source in the department.
According to Aleksandr Gusev, a member of the Supervisory Board of the National Medical Knowledge Base Association, Webiomed Development Director, the creation of the platform is an important step towards creating a market for AI services in the field of medicine in the Russian Federation. “With verified datasets, the machine learning needed for all of these projects will be possible. For example, we are developing a predictive analytics system that can predict the development of various diseases in a patient, and for its launch we would like to get access to datasets on cardiovascular diseases and diabetes, ”he says.
Experts note that the emergence of such services will affect the entire healthcare system. According to Pavel Roitberg, General Director of Digital Worlds, the competent use of AI services in healthcare can increase the speed of diagnostics, reduce the number of medical errors, and even carry out ultra-early diagnostics of patients' predisposition to pathological changes.
Anastasia Manuilova
Read also:
