Russia’s healthcare watchdog, Roszdravnadzor, is moving to automate the collection of performance data on artificial intelligence-based medical technologies, according to a report by the Vedomosti newspaper. The initiative targets clinical decision-support systems that analyze medical images—such as ultrasounds and MRIs—and electronic health records to help generate preliminary diagnoses.
A spokesperson for the agency told Vedomosti that developing the requirements “will help improve the monitoring of these products’ application,” adding that the draft document is currently under interagency discussion. Experts familiar with the matter noted that Roszdravnadzor currently lacks such a monitoring platform and will have to build it from the ground up.
This move mirrors the long-established tracking of adverse drug reactions but adapts it for the digital age. In the first case, the fact of an adverse event is identified by the person taking the drug; in the case of AI solutions, the system itself must collect information from devices about failures or incorrect operation, according to Boris Zingerman, director of the National Base of Medical Knowledge Association of AI Developers and Users.
According to Roszdravnadzor, 48 AI-powered medical devices are registered in Russia, 43 of which are domestic developments. These include:
· Botkin.AI: A platform for detecting pathologies in X-rays, CT, and MRI scans.
· Webiomed: A platform for predictive analytics, risk management, and clinical decision support.
· Third Opinion: A platform for processing MRI, CT, and X-ray data, digital blood and bone marrow smears, and analyzing video streams from medical institutions.
The planned data collection, while not containing direct patient identifiers, raises significant commercial sensitivity concerns because they allow an assessment of the quality of a specific device, according to Garnik Harutyunyan, Deputy Executive Director of MSU’s NTI on technologies of storage and the analysis of Big Data. Publishing information about errors can undermine trust in a particular medical device or organization, and information about errors and the effectiveness of software can reveal the product’s weak points, which is of commercial interest to competitors, clarified Igor Bederov, Director of the Investigations Department at T.Hunter.
The push for tighter oversight comes as AI gains a stronger foothold in Russian medicine, though public trust remains a hurdle. Surveys indicate that while Russians acknowledge AI can improve diagnostic accuracy, they remain reluctant to fully replace human doctors with algorithms.