Sber Medical Digital Diagnostic Center (MDDC)

Main articles:

• Artificial Intelligence in Medicine
• Decision Support Systems (DSS)
• Support systems of adoption of medical solutions (SPPR, CDS)

2021

Ecosystem of AI solutions for healthcare from SberMedAI: how to save resources in clinics and solve the problem of lack of doctors

What the health sector is facing today

1. Lack of an ecosystem approach for health digitalization.

Digitalization in the health sector accelerated in 2020, the driver of which was the pandemic. Interest in the introduction of innovative technologies in general and in artificial intelligence (AI), in particular, has increased sharply. AI in Russia has become universally used for the diagnosis of Covid-19. There are many disparate and not always verified IT solutions in order to put them into practice as effectively as possible, and therefore the need for an ecosystem approach in this area is ripe.

2. Polyclinics in the regions are equipped with modern equipment, but there are not enough highly specialized doctors to work.

The absence of doctors is a global trend that is only gaining momentum. According to the WHO forecast, by 2030 the world will face a shortage of doctors, 18 million doctors will be missing in the world.

At the same time, in our country the deficit is about 20%, especially in regions remote from the capital. However, the country's leadership sets the task of providing high-quality medical care in all territories, and modern technologies may well come to the rescue here. They can assemble a "digital consultation" for any clinic, regardless of its location.

Based on the needs of clinics in the Russian regions, the Medical Digital Diagnostic Center (MDDC) was created.

MDDC - An Artificial Intelligence Ecosystem for Healthcare

In August 2021, the company SberMedAI announced the launch of the Medical Digital Diagnostic Center (MDDC), a comprehensive solution for diagnosis using artificial intelligence and validation by a specialist doctor. On a single platform, more than 50 proven products and solutions of SberMedAI, as well as other ecosystem companies of Sber and partners, including: smart physician assistant "TOP-3," "," CT Lungs", "CT Stroke" as well as many others, were accumulated.

MDDC combines the development of SberMedAI, the Sber Artificial Intelligence Laboratory, the CoBrain-Analytics project of STI and Skoltech, as well as the solutions of ecosystem partners.

To create a comprehensive and most sought-after product, the market state for private healthcare was studied. Together with the ministries of health of the regions, the needs policlinics of doctors and doctors were identified, research was carried out on the basis of impersonal data patients in order to achieve maximum accuracy of artificial intelligence models.

Dozens of pilot projects were also carried out with both private and public clinics in many regions of the country. Doctors tested the products, gave feedback on their improvement. Today, the "TOP-3" model is installed in all adult clinics in Moscow, it is used by more than 3.5 thousand doctors, the "CT Lungs" service - in 25 medical institutions in more than 15 regions of the country, "CT Stroke" - in 6 regions.

The synergistic effect of combining services on the MDDC platform will allow clinics in the regions to gain seamless access to the spectrum of all healthcare solutions that use artificial intelligence.

Thanks to the launch of MDDC, Russian clinics, regardless of their location, will be able to use AI services this year. It is expected that they will be organically integrated into the medical information systems of medical and preventive institutions. The use of AI will increase the speed and quality of diagnostics.

Digital Consultation Platform

MDDC is designed to help doctors make diagnoses and make medical decisions based on data from primary admission, instrumental and laboratory diagnostics. The development can help solve existing problems of a shortage of doctors both in "narrow" specialties and in general in primary care, as well as increase the effectiveness and speed of diagnoses through accurate processing by artificial intelligence and additional verification of diagnosis by MDDC specialist physicians.

For example, if a clinic is faced with the need to promptly process a large number of CT examinations of patients with suspected COVID-19, then it needs to either increase the staff of radiologists or solve the problem by connecting the CT Lungs service. Thanks to the analysis of images, the radiologist can quickly and accurately determine the volume and proportion of pathology. There is no need for self-counting, which takes much longer than AI processing. With repeated tomographs, the service even indicates minor changes, so it is possible to track the dynamics of the disease and correct the treatment in a timely manner.

If a medical institution requires consultation on a difficult case in cardiology, then it needs to find a highly qualified doctor and agree with him on cooperation, which will take a long time. And a "digital consultation" will help solve the problem of "second opinion" literally during the day.

On the platform, clinics will be able to receive advice and assistance in a wide range of areas: therapy, cardiology, radiology, oncology, surgery, dental research. Institutions throughout the country can use the services of a "digital consultation."

How does it work

You can download digital documents through a web interface or a separate application for an automated workplace, or through programs already integrated into the clinic's medical information system (IIA).

Information is automatically routed and processed by AI algorithms, preliminary conclusions are submitted for verification to MDDC specialists and after that the recommendation for making a diagnosis returns to the clinic. At the same time, AI algorithms will take no more than 3 minutes to process. The time of verification by the MDDC consultant physician depends on the complexity of the case and can range from 15 minutes or more.

The MDDC platform provides a predictive selection of history specialists. And the advice of "digital doctors" takes place across a wide range. Within the framework of MDDC there are digital analogues of clinical-diagnostic, pathomorphological, histological laboratories. That is, as well as in "physical" laboratories, within the framework of MDDC, the results of KDL (clinical diagnostic laboratory studies) are processed, as well as the processing of images, sections of histological studies, etc. After processing by AI algorithms, the data are poisoned to the appropriate MDDC doctor for verification and formation of a conclusion.

Benefits for MDDC Implementation

The MDDC platform, implementing an ecosystem approach, provides new opportunities for medical institutions in Russia:

• Online access 24/7
• Rapid analysis of research using AI algorithms
• Verification of diagnoses by specialist doctors
• Wide range of specialized specialists
• Prioritizing the patient list
• Possibility of retrospective analysis of nosologies

Analysis of medical images - a solution that is being effectively applied now

Analysis of medical images is one of the most promising segments of AI use in healthcare at the moment. It allows you to significantly facilitate the work of the doctor, speed up the processing of data, quickly form the necessary "hints," thereby improving the quality of diagnosis and medical care.

MDDC is a platform for working with medical images with the following characteristics:

• Full functionality for use as a regional radiological information system (RRIS) or LCP RIS
• Transfer and storage of medical images, including DICOM 3.0 format
• Automate analysis of medical imaging data, including AI algorithms for analysis of medical images
• Automation of data exchange with external information systems
• Remote data access ("second" opinion)
• Provision of advice on radiation diagnostics to medical organizations in case of necessity and absence of specialists in the staff

Among the advantages of the MDDC platform are the ability to centrally store and visualize medical data, effectively manage the resources of the radiation diagnostics service, and reduce the cost of ownership of the medical image archive.

Connection to the system of Primorsky and Khabarovsk Territories, Amur and Sakhalin Regions

Primorsky Krai became one of the first regions of Russia whose medical information system connected to MDDC. This was announced on September 2, 2021 in the company SberMedAI.

As of September, MDDC combines on a single platform more than 50 products and solutions of SberMedAI, as well as other companies of the Sber ecosystem and partners, including a smart physician assistant "TOP-3," "CT Lungs", "CT Stroke", etc.

Doctors of clinics connected to the platform online can send to the "digital consultation" data of primary reception, instrumental and laboratory diagnostics. You can download digital documents through a web interface, or a separate application for an automated workplace, or through programs already integrated into (medical information system IIAs) clinics. At the same time algorithms AI, it will take no more than three minutes to process. But the time of verification by the MDDC consultant doctor depends on the complexity of the case and can be from 15 minutes or more, they told SberMedAI.

In the health care of the Far East, a number of services with artificial intelligence are already being used, and now the region is one of the first in Russia to connect to the MDDC platform, which combines all these developments using AI. This ensures access to the competencies of doctors in narrow specialties, as well as increases the effectiveness and speed of diagnoses, regardless of long distances, "said Igal Zak, CEO of SberMedAI.

In the Far East, developments with AI, presented on the platform of the Medical Digital Diagnostic Center, have already been introduced in ten treatment and preventive institutions. On Sakhalin, in the Khabarovsk and Amur Regions, the CT Lung service, developed in collaboration with the Sber Artificial Intelligence Laboratory, is used to combat COVID-19, helping doctors track pathological changes, as well as segmentation of areas of affected lung tissue indicating the volume of damage.

CT Stroke is used by doctors in the Khabarovsk Territory and the Sakhalin Region. Based on AI algorithms, it automatically marks CT images, allows you to quickly and accurately assess stroke damage both by quantifying the volume of ischemia and stroke type. Also in the Khabarovsk Territory an artificial intelligence model helps to classify mammographic studies with the definition of a BI-RADS group with the ability to automatically highlight suspicious lesions.

Medical Digital Diagnostic Center Concept

On June 3, 2021, SberMedAI introduced the concept of a medical digital diagnostic center (MVDC or MDDC), which will operate using AI technologies.

The virtual diagnostic center will be designed for medical and preventive institutions in Russia and embedded in medical information systems and workstations of doctors.

Incoming from medical and preventive institutions (LPUs) data will be processed by algorithms AI, preliminary conclusions will be submitted for verification to specialist doctors, after which the recommendation for making a diagnosis will return to the LPU.

The service is created to support medical decision-making based on data primary reception, instrumental and laboratory diagnostics. With its help Russian clinics, regardless of their location, they will be able to increase the availability of high-quality medical care.

The platform provides the possibility of predictive selection of specialists with a history. They are provided with counselling in a wide range of areas such as therapy, cardiology, radiology, oncology, surgery, dental research.

SberMedAI also presented AI services for analyzing CT images: "CT Stroke" and "CT Lungs". Services are designed primarily for radiologists. The use of AI allows you to significantly facilitate the work of the doctor, speed up the processing of data, quickly generate the necessary "hints," thereby improving the quality of diagnosis.