NPCC DiT DZM: Unified Radiological Information Service (ERIS)

The main articles are:

• Radiology
• Artificial intelligence in radiology
• Artificial Intelligence in Medicine
• Unified Medical Information and Analytical System (EMIAS)

The Unified Radiological Information Service (ERIS) EMIAS is an information system that combines high-tech medical equipment, workplaces of X-rays, radiologists, an advisory service, and a single archive of diagnostic images.

2024

Domestic AI service surpassed foreign ones in terms of accuracy of detecting nodes in the lungs

Moscow scientists assessed the parameters of diagnostic accuracy AI- services of radiation diagnostics. Five took part in the study: three of algorithms them were developed in, the Russia other two in and. India South Korea The domestic AI service has become the best in terms of accuracy of determining pulmonary nodes. data medical that Zdrav.Expert Representatives shared these with the portal on October 16, 2024. Diagnostic and Telemedicine Center Department of Health of Moscow

"We prepared a unique set data consisting of 100 radiographic examinations: 50 with pulmonary nodes and 50 without pulmonary nodes. A distinctive feature of this data set: out of 50 radiographs with pulmonary nodes 25 - contain pulmonary nodes that were unambiguously determined by the radiologist. In the other 25 - the doctor doubted. However, the presence of pulmonary nodes in all 50 studies was confirmed using, computed tomography- said medical information scientists telemedicine Kirill Arzamasov, head of the department, radiomics and radiogenomics of the Center for Diagnostics and DZM.

As he explained, the assessment of AI services was carried out in 3 stages. At each stage, the results of AI services were compared with the reference markup confirmed by CT examination. The study showed that the domestic AI service surpassed foreign ones in all indicators of diagnostic accuracy. The dataset is publicly available on the^[1] website] and can be used to assess the quality of AI services by the developers themselves.

"In a recent study involving Russian and foreign artificial intelligence services, our domestic AI service demonstrated the highest accuracy in detecting pulmonary nodes, surpassing similar solutions from India and South Korea. This is an important step in the development of technologies to improve the quality of medical care, "said the chief freelance specialist in radiation and instrumental diagnostics of the Department health care Moscow, director of the Moscow Center for Diagnostics and Telemedicine. Yuri Vasiliev

According to him, radiologists can use more than 50 artificial intelligence services to describe medical images. With the help of neural networks in Moscow, more than 13 million radiation studies have already been analyzed. The AI services market offers more and more solutions that simplify the work of a doctor, while maintaining the quality of its implementation.

We strive to provide doctors with tools that not only simplify their work, but also improve the quality of medical care. To do this, we created a maturity matrix - a tool that shows the results of services and allows you to evaluate them. From the end of 2022 to October 2024, this AI service occupies a leading position in the direction of "Chest X-ray," "added Yuri Vasiliev.

The AI service is designed to automatically analyze chest X-rays. As of mid-October 2024, it identifies 14 radiological signs of various pathologies. The service also calculates the cardiothoracic index and forms a complete draft study protocol.

The study is published in the Chinese scientific journal Quantitative Imaging in Medicine and Surgery.

An online service has been launched in Moscow to track the download of medical equipment in clinics

In October 2024, a new online tool began work in Moscow, allowing real-time tracking of the download of medical equipment in city polyclinics and hospitals. The service, developed by the diagnostic and telemedicine center of the Moscow Department of Health, is designed to optimize the use of diagnostic equipment and increase the availability of medical services for patients.

According to reports, TASS the new information panel provides access to data on the download of all digital diagnostic equipment connected to the Unified Radiological Information Service. EMIAS The system covers 155 medical organizations subordinate to the Department. health care Moscow

Online service for tracking the download of medical equipment launched in Moscow

Yuri Vasiliev, chief freelance specialist in radiation and instrumental diagnostics in Moscow, director of the center for diagnostics and telemedicine, stressed that the new service will allow specialists in clinics and hospitals to independently analyze data on the loading of both their own devices and equipment of other medical institutions. This will make it possible to plan work more efficiently, redistribute the load on devices and patient flows, as well as make informed management decisions.

The online service has wide functionality and allows you to configure various filters for the medical organization, type of research, administrative district and type of device. Users can check information on more than 20 parameters, including the percentage of equipment load, the number of studies performed, the shift of work and the technical characteristics of the devices.

The system allows you to view data in table format with information on diagnostic devices for different time periods. An important feature of the service is the regular update of data, which ensures the relevance of the information provided.^[2]

The first AI service has been launched in Moscow to identify and assess the displacement of vertebral bodies

A service with artificial intelligence (AI) has become available to Moscow doctors, which helps to identify and assess the displacement of vertebral bodies on radiographs. This was announced to the medical portal Zdrav.Expert on September 9, 2024 by representatives of the Center for Diagnostics and Telemedicine of the Moscow Department of Health. According to them, "this is a new direction in the development of such technologies." In total, more than 50 AI services are available to radiologists in the capital, who find signs of 38 pathologies on radiation studies, said Anastasia Rakova, Deputy Mayor of Moscow for Social Development.

"The displacement of vertebral bodies is one of the possible causes of back and low back pain, which can significantly reduce the quality of life of Muscovites. It is important to diagnose this condition in time and accurately to avoid complications and prescribe the correct treatment. The new AI-based service we have implemented allows doctors not to miss the bias when describing radiographs of the spine, automatically taking the necessary measurements. This speeds up the preparation of the conclusion and helps to start treatment as early as possible, "said Anastasia Rakova.

Displacement of vertebral bodies is a condition in which the vertebrae emerge from their normal position, often leading to pain in the lower back and other parts of the back. This can be the result of trauma, degenerative changes, or diseases such as osteoporosis. Accurate diagnosis is important for prescribing the right and timely treatment, and AI services can play an important role here. Neural networks are able to automatically analyze radiographs of the spine, detect signs of displacement and instantly take measurements, which helps doctors diagnose pathology faster and more accurately. This reduces the time required to prepare a conclusion and allows you to prescribe treatment more quickly.

Neural networks help to find signs of various pathologies on mammograms, CT, MRI and X-rays, including lung cancer, pneumonia, osteoporosis, aortic aneurysm, coronary heart disease, stroke, pulmonary hypertension, hydrothorax and other diseases. The accuracy of diagnostics of artificial intelligence services is comparable to the accuracy of medical research, noted in the Center for Diagnostics and Telemedicine.

The quality of the algorithms is monitored by specialists from our Center for Diagnostics and Telemedicine of the Moscow Department of Health. Only services that have successfully passed all tests are allowed to work with real research, said Yuri Vasiliev, chief freelance specialist in radiation and instrumental diagnostics of the Moscow Department of Health, director of the Moscow Center for Diagnostics and Telemedicine.

"At the same time, when they go to streaming, we monitor the technological and clinical performance of services, assess the accuracy of algorithms, analyze the identified discrepancies between the conclusion of the service and the doctor. This analytics and feedback from scientists and radiologists is provided to developers. They use this information to further learn and improve their models. Thus, we ensure the reliable operation of neural networks, "the specialist explained.

More than 12 million images were processed using neural networks, signs of 37 different diseases are detected

The capital has been using computer vision technologies in medicine for the fifth year. During this time, neural networks processed more than 12 million images and learned to detect lung disease, heart disease and signs of oncology. Artificial intelligence (AI) analyzes fluorograms, mammograms, radiographs, magnetic resonance and computer tomograms, finds signs of pathologies and takes the necessary measurements. On May 31, 2024, the Moscow Department of Information Technology and the Moscow Department of Health told about how smart algorithms help doctors.

As reported, computer vision is one of the varieties of artificial intelligence that is used to analyze images. After the study, the digital medical image is uploaded to the unified medical information and analytical system (EMIAS) and becomes available to the neural network. The service returns the results of the analysis with color markings of areas of possible pathologies, the necessary descriptions and measurements. The radiologist sees both the processed image and the original one. Such a process helps optimize the accuracy of diagnoses and saves doctors time.

When developing neural networks for the analysis of radiation studies, a deep learning method is used, which allows you to achieve optimal accuracy in diagnostics.

The neural network learns from large amounts of data. In the process of analyzing the image, the model applies various filters and algorithms to highlight key areas in the image and highlight potential anomalies. The system then uses ranking algorithms to determine the probability of a particular disease. At the same time, the final decision on the diagnosis always remains with the specialist. This provides medical control. The neural network in this case acts as an assistant doctor, allowing you to optimize diagnostics, reducing the time for describing research. noted in the Department of Information Technologies of the city of Moscow

As of May 2024 hospitals policlinics , more than 50 specialized services based on neural networks are available to doctors in the capital. Computer algorithms features of 37 different diseases are detected. Among them are cancer of the lung, osteoporosis spine, aneurysm of the thoracic region,, and aortas coronary heart disease pulmonary hypertension hydrothorax, emphysema lung, stroke, adrenal gland formation.

The implementation of the project made it possible to create and develop a market for artificial intelligence services in radiation diagnostics. As of May 2024, the competitive environment of AI service developers is supported: the maturity matrix of solutions developed by the participants in the experiment is openly published. told Yuri Vasiliev, chief radiologist of Moscow, director of the Center for Diagnostics and Telemedicine of the Moscow Department of Health

In addition, computer vision sees signs of lung pathologies on radiography and fluorography, flat-footedness on radiography of the feet, arthrosis of the knee scoliosis joint and spine on radiography musculoskeletal system, as well as protrusion and herniation of intervertebral discs, spinal canal stenosis on magnetic resonance imaging of the lumbosacral spine, breast cancer on mammography.

Thanks to the introduction of smart algorithms into medical practice, doctors can perform diagnostics much faster. For example, the time to describe preventive mammograms has been reduced by eight times.

In February 2024, Moscow opened access to a special platform "MosMedII" with artificial intelligence services for medical organizations from all over the country. Using the platform, doctors will be able to obtain the results of automatic analysis of radiation studies performed by smart algorithms. Applications for connection have already been received from more than 25 regions, including the republics of Bashkortostan and Dagestan, St. Petersburg, Moscow, Penza and Kemerovo regions, Krasnoyarsk Territory, Yamalo-Nenets Autonomous Okrug. Both individual organizations and all medical institutions in the region can access the platform. In addition, the Moscow digital library has recently been replenished with updated sets of data on radiation diagnostics.

Among them are seven open datasets for mammography, chest X-ray, fluorography and computed tomography. A total of 68 have been created to evaluate and test neural networks. Any developer can check the abilities of the neural network to look for signs of pathologies in medical images. In 2023, several complex services with artificial intelligence were introduced. They are able to detect signs of many pathologies on a single computed tomogram.

In the capital, the analysis of images by artificial intelligence is included in the compulsory health insurance program. Moscow was the first in the country to introduce a special tariff for the analysis of radiation research using neural networks. For citizens, such research will remain free.

Artificial intelligence services are integrated into a single radiological information service of a single medical information and analytical system of the capital. Thus, radiologists of all medical institutions of the city connected to EMIAS can use modern technologies. The project is being implemented on the basis of the Center for Diagnostics and Telemedicine of the Moscow Department of Health together with the Department of Information Technologies of the city of Moscow.

Addition of 7 open datasets for mammography, chest X-ray

The Moscow library of data sets for AI has been replenished with new data sets. This was announced on May 3, 2024 by the press service of the Center for Diagnostics and Telemedicine DZM.

Among the changes are 7 open datasets for mammography, X-ray analysis chest organs, fluorography and. computed tomography Thus, as of May 2024, the employees of the Center for Diagnostics telemedicine Moscow Department of Health and created more than 350 data sets for assessing and testing work, neuronets among which 68 are available to all network users. Internet Most are designed for self-testing. algorithms

Published open-label datasets can be used in the self-testing of neural networks to search for signs of pathologies of the brain, chest organs and other anatomical areas. The digital library appeared in 2022, is presented in all the main modalities of radiation diagnostics and is gradually being replenished with new data sets. All open datacets are published on the platform mosmed.ai.

At the beginning of 2024, specialists from the Center for Diagnostics and Telemedicine health care of the City Department Moscow created the first open set of computer brain tomography data. It includes 800 CT studies, additional clinical and technical. information

2023

In Moscow, for the first time, artificial intelligence began to detect ENT diseases on X-rays

To Moscow neuronets For the first time, they began to recognize the signs of ENT diseases in the x-ray pictures. Smart algorithms are able to detect a serious pathology - sinusitis from the results of sinus radiography. This is the most common study to detect diseases of the nasal sinuses. In 2022 Russia , more than three million such surveys were conducted in. A service based on () artificial intelligence AI will help doctors get preliminary results immediately after the procedure. This was Center for Diagnostics and Telemedicine (NPCC DiT DZM) announced on November 14, 2023.

Radiation diagnostics are developing faster than others health care in the field of digital services and neural networks. This year alone, Moscow doctors have services in nine areas of artificial intelligence at once. In Moscow, artificial intelligence has learned to determine the signs of ENT diseases on X-rays of patients. This is already the 28th clinical direction in which algorithms they can recognize pathologies. The preliminary result of the study processing will be available within two to three minutes. The neural network significantly facilitates the task of the doctor and saves time that can be spent on the patient, - said the Anastasia Rakova Deputy Mayor of Moscow for Social Development.

The AI-based service marks areas of possible pathologies with color hints on the medical image, makes automatic measurements that are important for the doctor when making a diagnosis, and compiles a description. The final decision is made by the doctor, but the use of technology significantly speeds up and increases the accuracy of work.

As of November 2023, more than 50 artificial intelligence services in 28 clinical areas are operating in the capital. With their help, doctors analyzed over 10.5 million studies.

Correctly describing X-rays in sinusitis is often a non-trivial task. The front of a person's skull is quite complicated, so in an X-ray, different bone structures are superimposed on each other. Against this background, it is sometimes difficult to assess inflammatory changes in the soft tissues of the sinuses and other manifestations of pathology, for example, the level of fluid that the sinus can be filled with, or even violations of the integrity of the bone walls. Artificial intelligence in such cases may turn out to be a useful tool that will draw the doctor's attention to the thickening of the mucous membranes and calculate the percentage of sinus filling with pathological contents, "said Yuri Vasiliev, chief radiologist of Moscow, director of the Center for Diagnostics and Telemedicine.

The use of computer vision technologies in medicine reduces the time to interpret diagnostic studies, and also provides medical staff with information to make more accurate diagnoses and prescribe effective treatment. The introduction of such technologies in healthcare allows the creation and development of a market for support systems for medical solutions in radiation diagnostics.

Moscow doctors can already use 5 neural networks to help in determining diseases of the chest organs

Moscow doctors can already use 5 neural networks to help in determining diseases of the chest organs. This was announced on October 31, 2023 by the Center for Diagnostics and Telemedicine (NPCC DiT DZM).

The neural network is capable of recognizing signs of seven pathologies, including lung cancer, pneumonia, spinal osteoporosis, aortic aneurysm, pulmonary hypertension, hydrothorax, coronary heart disease by the degree of calcification of the coronary arteries and the volume of paracardial fat. Now, when describing radiation images, radiologists can use one of six complex solutions.

A new service based on artificial intelligence has begun work in Moscow to comprehensively determine seven diseases of the chest organs. Now doctors, when analyzing such studies, can choose from the five available neural networks the most convenient and effective. Development requirements are constantly growing, which contributes to their development. Last year, only one comprehensive service worked, and now six software solutions are available to doctors. This is a qualitatively different level of offered assistance to doctors from developers of artificial intelligence, - said Anastasia Rakova, Deputy Mayor of Moscow for Social Development.

In the medical image, the neural network highlights deviations in color, makes the necessary measurements, and also draws up a text description. Algorithms work as part of a scientific experiment to introduce computer vision technologies into radiation diagnostics.

The fifth service began to analyze CT studies of the abdominal cavity of patients. It helps to identify signs of adrenal neoplasms, compression fracture of vertebral bodies, abdominal aortic aneurysm, urolithiasis, liver and kidney neoplasms. In total, more than 10.5 million medical images were analyzed using artificial intelligence.

In the experiment, more and more complex solutions appear that are able to determine several pathologies at once, instead of monoservices to identify only one pathology. Its organizers set the vector for the creation and development of intellectual developments. Thus, the transition to comprehensive solutions was announced at the end of 2021, and today six software products are available to doctors. Proposals from developers are becoming more and more competitive, and doctors, in turn, get the opportunity to choose the most convenient services for themselves, "said Yuri Vasiliev, chief radiologist of Moscow, director of the Center for Diagnostics and Telemedicine.

200 thousand mammograms paid for compulsory medical insurance were processed

In 2023, Moscow became the first region in which mammographic studies using artificial intelligence are paid from the compulsory medical insurance fund. During this time, neural networks have already processed 200 thousand images. This was announced on October 11, 2023 by the deputy head of the Moscow Health Department Ilya Tyrov.

Mammography is the most important study for the prevention of breast cancer, as it allows you to identify pathology in the early stages. In Moscow, descriptions of preventive mammograms by artificial intelligence have been carried out within the framework of the compulsory medical insurance system for 9 months. As of October 2023, the algorithms processed 200,000 images. The use of this technology has already shown good results: we see that the accuracy of neural networks is comparable to the work of a doctor, and thanks to smart algorithms, the time for describing images has been reduced by 8 times. Double reading of mammographic images makes the assessment independent and high-quality, - said Ilya Tyrov.

The inclusion of analysis of radiation images by neural networks in the compulsory medical insurance tariff allows you to introduce such technologies into practical healthcare and not depend on grant support. For residents of the city, these studies are still free, and the use of smart algorithms allows you to speed up diagnostics and increase its accuracy.

Moscow specialists conducted a study and compared two scenarios for analyzing screening mammographic images: two radiologists and a specialist and artificial intelligence. Experts have found that smart algorithms speed up the description of mammograms by more than 8 times. Thus, the use of artificial intelligence in the analysis of mammograms allows to reduce the burden on radiologists and increase the speed of diagnosis of breast cancer.

Algorithms work as part of a scientific experiment to introduce computer vision technologies into radiation diagnostics.

For the work of artificial intelligence, solutions that have registration certificates are selected within the framework of compulsory medical insurance. medical device We set a high threshold for services to enter at the experimental stage: developers create solutions, then they go through several stages of testing and, only in case of successful results, the service is allowed to work with real research. Thus health care Moscow , proven artificial intelligence services in popular areas get into the system, "said the chief radiologist of Moscow, director. Center for Diagnostics and Telemedicine of the Metropolitan Department of Health Yuri Vasiliev

The fifth comprehensive AI service can reveal six pathologies at once in one picture

B To Moscow earned the fifth comprehensive service artificial intelligence for diagnosing diseases. Algorithm is able to find computed tomography signs of six pathologies on one image at once Moscow , doctors have not previously used them to identify them. neuronets These include urolithiasis, as well as neoplasms and. liver kidneys This was announced on October 4, 2023, Anastasia Rakova by the Deputy Mayor of Moscow for Social Development.

We continue to expand the Moscow experiment on implementation computer vision beam in diagnostics. Another, already fifth, comprehensive service has begun to work, artificial intelligence which finds computer signs of six diseases in the tomography images of the abdominal organs. Neural networks analyze the image to detect adrenal neoplasms, compression fracture of bodies, vertebras aneurysms abdominal region, as well aortas as in three directions - diagnosis of urolithiasis, liver and kidney neoplasms. The key feature of the integrated service is that artificial intelligence will draw the attention of the doctor even to those pathologies that are not the subject of examination, Anastasia Rakova told.

She added that the comprehensive service is the first to detect six abdominal diseases. The rest are aimed at diagnosing pathologies of the chest organs. One of them is able to find signs of 10 pathologies on a computed tomogram. Among them are adrenal neoplasms, lung cancer, pulmonary hypertension, pneumonia, osteoporosis and wedge deformation of vertebral bodies, aneurysm of the thoracic aorta, hydrothorax, pulmonary emphysema, coronary heart disease by the degree of calcification of the coronary arteries and paracardial fat volume. Two complex services determine the signs of seven pathologies, and one - eight.

The first stage when opening the direction of the experiment, we determine its demand in the conditions state health care of the city system. We analyze the number of medical Moscow Department of Health studies conducted per year in organizations, determine the most frequently performed of them and the most detected pathologies. Next, we announce on the website project the directions for which we are ready to take part in the experiment intellectual developments, publish diagnostic and functional requirements for them. The developers create solutions, then we create them, and in we test case of successful test results, the service is allowed to work with real research that is carried out in Moscow medical organizations. Thus, only proven artificial intelligence services in popular areas become available to radiologists, told Yuri Vasiliev, chief radiologist of Moscow, director of the Center for Diagnostics and Telemedicine of the Moscow Department of Health.

Neural networks are used to process images of radiation studies obtained during fluorography, mammography, radiography, computer and magnetic resonance imaging. As of October 2023, the project has more than 50 services based on artificial intelligence in 26 areas of research help radiologists in recognizing various diseases.

The algorithms work as part of an experiment to introduce computer vision technologies into healthcare. The project implements the complex of social development of Moscow together with the Moscow Department of Information Technologies on the basis of the Center for Diagnostics and Telemedicine, an advanced scientific and practical organization in the structure of the Moscow Department of Health, founded in 1996.

Artificial intelligence processed 1 million mammograms of metropolitan patients in 3.5 years

In Moscow, for 3.5 years, 1 million mammographic studies were analyzed using artificial intelligence services. Neural networks have been operating in the capital since 2020 as part of an experiment on the use of computer vision technologies in healthcare. Radiologists for October 2023 use 4 artificial intelligence services to analyze mammographic examinations. This was announced on October 2, 2023 by the deputy head of the Moscow Department of Health Ilya Tyrov.

Mammography is a type of research in which artificial intelligence has become an indispensable doctor's assistant. Moscow in 2023 introduced a special tariff for the analysis of preventive mammograms using neural networks under the compulsory health insurance program. Screening for women involves a large stream of studies described by a radiologist. Partial automation of this process helps to cope with volume quickly and efficiently. The final decision remains with the doctor, - said Ilya Tyrov, deputy head of the Moscow Health Department.

He also added that, according to scientists, artificial intelligence services reduce the time for describing mammography studies by more than eight times. In addition, neural networks have diagnostic accuracy at a level comparable to the work of a radiologist.

Our task is to ensure the accuracy of diagnostics using neural networks, so we constantly monitor the quality of work of already functioning algorithms so that doctors can only apply reliable solutions, "said Yuri Vasiliev, chief radiologist of Moscow, director of the Center for Diagnostics and Telemedicine.

The fourth complex of AI service for the diagnosis of diseases appeared in Moscow

The fourth comprehensive artificial intelligence service has appeared in Moscow, which can simultaneously detect several diseases on a computed tomogram of the chest organs. The algorithms work within the framework of the Moscow experiment on the use of computer vision technologies in radiation diagnostics. This is the fourth comprehensive solution in the domestic market of algorithm developers. This was announced on September 27, 2023 by the acting Deputy Mayor of Moscow for Social Development Anastasia Rakova.

Another neural network has appeared in the capital to comprehensively identify signs of chest diseases. The emergence of such solutions is the result of the development of the domestic market for intellectual developments in the field of medicine, which is due to the qualitative expansion of the Moscow experiment. Thanks to the automation of routine tasks, doctors have more time to analyze patient studies. At the same time, the team of metropolitan specialists continues to search for optimal scenarios for the use of high-tech complex solutions in practical activities. As of September 2023, in Moscow, about 50 algorithms in 23 clinical areas analyze research in real time, four of which are complex AI services, - said Anastasia Rakova.

According to her, in one study of the chest, the neural network is able to recognize the signs of 8 pathologies: lung cancer, pneumonia, spinal osteoporosis, aortic aneurysm, pulmonary hypertension, hydrothorax, coronary heart disease by the degree of calcification of the coronary arteries and the volume of paracardial fat.

The first comprehensive service of artificial intelligence began to help metropolitan radiologists a year ago. On average, he processes about 20 thousand radiation studies per month. On one computed tomogram, it is able to simultaneously recognize signs of up to 10 diseases. The second comprehensive solution appeared in the spring, the third - in the summer of 2023. In total, over the period of the experiment, artificial intelligence services analyzed more than 10.5 million medical images.

In the medical image, the algorithms highlight the abnormal areas with color hints and make the necessary measurements. Both visual cues and a textual description are provided to the physician.

For 8 months, more than 6 million CT, MRI and other radiation studies were carried out in Moscow

Since the beginning of 2023, more than six million radiation studies using digital heavy medical equipment have been carried out in the capital. This was announced on September 22, 2023 by Ilya Tyrov, Acting Deputy Head of the Moscow Department of Health.

Moscow has been engaged in the digitalization of the health care system for more than 10 years, and one of the important stages of this work was the development of a single radiological information service (ERIS) of a single medical information and analytical system (EMIAS). It made it possible to centralize the radiation diagnostic service and develop a model for describing radiation studies. As a result, the Moscow Reference Center for Radiation Diagnostics was formed. Since the beginning of 2023, more than six million CT scans , MRI and other radiation studies have been performed in Moscow, over three million of which were analyzed at the reference center, "said Ilya Tyrov.

Today, radiological studies that are carried out in adult polyclinics are analyzed in a single round-the-clock reference center. He works as part of the Center for Diagnostics and Telemedicine of the Metropolitan Department of Health. Any resident of Moscow can receive the opinion of a highly qualified narrow-profile expert by performing the procedure in his district clinic and without leaving the other end of the city. The medical image automatically enters ERIS EMIAS, from where it is taken for description by a radiologist. Within hours, the completed description and study appear in the patient's electronic medical record.

As of September 2023, almost all radiologists of adult polyclinics in the capital, which is more than 300 doctors, work at the Moscow Reference Center, describing radiation studies of patients around the clock. Centralization of radiation diagnostics made it possible to improve the quality of work of radiologists and reduce the time for preparing an opinion for a patient. Over the entire existence of the unit, more than eight million descriptions of radiation studies have been carried out: computer and magnetic resonance imaging, mammograms and X-ray images. And every year quantitative and qualitative indicators are growing: since the beginning of 2023, more than three million descriptions have already been processed, which is almost twice as much as in the same period last year, - explained the chief radiologist of Moscow, director of the Center for Diagnostics and Telemedicine of the Moscow Department of Health Yuri Vasiliev.

Now artificial intelligence helps Moscow doctors identify signs of fractured ribs on CT

In the Moscow Art Intelligence, on CT studies of the chest organs, he learned to determine the fracture of the ribs. This tool will help doctors in describing complex diagnostic cases, for example, in combined pathology or severe injuries. This direction will be included in the complex service of artificial intelligence, that is, algorithms will be able to determine in one picture the signs of many pathologies at once. This was announced on August 22, 2023 by the Deputy Mayor of Moscow for Social Development Anastasia Rakova.

{{quote 'A solution based on artificial intelligence technologies is now available to Moscow radiologists, which allows you to quickly and accurately find rib fractures, as well as assess the integrity of the bone structures of the chest on CT. The service will be especially useful in diagnosing complex cases - algorithms automate routine processes, make the necessary measurements and pay the doctor's attention to subtle signs of pathologies, that is, save a specialist's time and improve the quality of diagnostics. In addition, the direction will later become part of a comprehensive artificial intelligence service, which will allow determining the signs of several pathologies at once within the framework of one study. For example, a patient is examined in the chest with suspected rib fracture, and the neural network will additionally analyze the picture for possible pathologies of the lungs or heart, - said Anastasia Rakova. }}

The deputy mayor added that as of August 2023, the project already has more than 50 artificial intelligence services in 23 clinical areas, in total, doctors analyzed about 10.5 million medical images with their help.

The use of computer vision technologies in medicine reduces the time of diagnosis, and also provides medical staff with information to make more accurate diagnoses and prescribe more effective treatment. Thanks to artificial intelligence, the availability of medical care is also increased.

We are introducing many digital services that facilitate and speed up the work of doctors. This service of artificial intelligence will allow a comprehensive assessment of the state of the chest organs within the framework of one study - both the pathology of the lungs and mediastinum, and damage to bone structures. An undoubted plus is the textual description of the study by artificial intelligence, the doctor may not spend a lot of time drawing up a protocol, but simply use the necessary fragments of the text, "said Yuri Vasiliev, chief freelance specialist in radiation and instrumental diagnostics in Moscow, director of the Moscow Center for Diagnostics and Telemedicine of the City Health Department.

Algorithms mark areas of possible pathologies with color hints on the medical image. The neural network automatically makes measurements that are important to the doctor at diagnosis and compiles a description. Computer vision technologies allow the detection of pathologies at an early stage, when they can be invisible to the human eye.

Development of principles for standardization of datacets for AI training in medicine

Moscow scientists have developed principles for standardizing datacets for training artificial intelligence in medicine.

The developers train neural networks to independently find pathologies in radiation studies of patients using datacets - sets of impersonal radiological images with signs of certain diseases. This was announced on July 5, 2023 by Anastasia Rakova, Deputy Mayor of Moscow for Social Development.

We are not only engaged in the introduction of artificial intelligence technologies in the capital's healthcare, but also pay great attention to the scientific side of the issue. Moscow radiologists have proposed an original technology for structuring and systematizing the management of medical datasets for training and testing artificial intelligence. This will standardize information about data sets for machine learning. Creating a registry makes it possible to quickly form visual data libraries, allowing a wide range of researchers, developers and companies to choose datasets for their tasks. The creation of such an original technology for the world market ensures widespread use and contributes to the rapid development and implementation of services based on artificial intelligence, - said Anastasia Rakova.

When developing algorithms for maintaining the register, the main features were taken into account, including information management, data quality, confidentiality and security.

Scientific the study took place in four phases, which included an analysis of all available resources with datacets. A large library, including with open sets, data was developed by experts from our Center. As of July 2023, about 200 datacets, some of which are available to any network user, are posted on the website of the Moscow experiment. A great contribution to the creation of the registry was made on the basis of an analysis of the data obtained during the experiment on the implementation computer vision of radiation diagnostics. Our scientific developments are now available to a large circle, they are published in a peer-reviewed scientific journal, - said the deputy director for scientific work of the Center for Diagnostics and Telemedicine DZM. Anton Vladzimirsky

Ability to simultaneously detect up to 10 pathologies on a computed tomogram

Moscow has expanded the capabilities of a comprehensive artificial intelligence service. The neural network can now simultaneously detect up to 10 pathologies on a computed tomogram. The diseases that algorithms help doctors determine on CT images of the chest included adrenal glands. The Center for Diagnostics and Telemedicine announced this on June 27, 2023.

The development of complex solutions based on artificial intelligence is one of the main directions of the experiment on the introduction of computer vision technologies into radiation diagnostics in 2023. We already have two complex tools that can reveal several pathologies in one picture at once. Recently, the neural network has learned to recognize adrenal gland formations on a chest CT scan. Thus, one service already determines up to 10 deviations in one picture. This makes it possible to diagnose pathologies in patients who have seen a doctor to treat other diseases. In total, since the beginning of the work of the complex service of artificial intelligence, he helped doctors analyze more than 270 thousand medical images, - said Anastasia Rakova.

In addition to adrenal gland formations artificial intelligence , it can simultaneously find computer signs of,, lung cancer COVID-19 osteoporosis spine, aneurysms thoracic aorta, pulmonary hypertension hydrothorax, pulmonary emphysema, ischemic disease on the tomogram by hearts the degree of calcification of the coronary arteries and the volume of paracardial fat.

Algorithms mark areas of possible pathologies with color hints on the medical image. The neural network automatically makes measurements that are important to the doctor at diagnosis and compiles a description. Computer vision technologies allow the detection of pathologies at an early stage, when they can be invisible to the human eye.

The second comprehensive artificial intelligence service for diagnosing 7 diseases at once has been launched in Moscow

A To Moscow second comprehensive service has appeared in, artificial intelligence which computer can simultaneously detect seven diseases on a tomogram. In one study of the chest neuronet , it is able to recognize the signs, lung cancer COVID-19 osteoporosis spine, aortic aneurysms, pulmonary hypertension hydrothorax, by coronary heart disease the degree of calcification of the coronary arteries. This is the second comprehensive solution from domestic developers - doctors will be able to choose which service is convenient to use. This was announced on June 8, 2023 by the Deputy Mayor for Moscow Social Development. Anastasia Rakova

For three years now, we have been actively introducing artificial intelligence technologies into the capital's healthcare. One of the key areas is computer vision in radiation diagnostics. Since the start of the Moscow experiment on the use of smart neural network algorithms, more than 9 million radiation studies have been analyzed. We have already launched the second comprehensive artificial intelligence service, which can simultaneously detect 7 diseases on CT. reported by Anastasia Rakova

As reported, the first comprehensive artificial intelligence service began to help metropolitan doctors a year ago. On average, he processes about 25 thousand radiation studies per month. On one computed tomogram, he is able to simultaneously recognize the signs of 9 diseases: lung cancer, COVID-19, spinal osteoporosis, thoracic aneurysm aorta, pulmonary hypertension, hydrothorax, coronary heart disease by coronary artery calcification, coronary heart disease by paracardial fat volume, and pulmonary emphysema.

To participate in the experiment, you need to go through several stages - submit an application, integrate into the Unified EMIAS Radiological Information Service, undergo testing, testing and go into trial operation. At each stage, technological and clinical performance indicators are monitored. We assess the accuracy of algorithms, analyze together with the developers the work of the service and the doctor, transmit feedback from radiologists to them. noted director Yuri Vasiliev, Center for Diagnostics and Telemedicine of the Moscow Department of Health, chief freelance specialist in radiation and instrumental diagnostics of the capital

In the medical image, the algorithms highlight the abnormal areas with color hints, make the necessary measurements for the doctor, and also compile a text description. For June 2023, algorithms are available radiologists of about 150 medical institutions, including children's. The algorithms work within the framework of a scientific experiment on the introduction of computer vision technologies into radiation diagnostics, which is implemented by the Moscow Social Development Complex on the basis of the Center for Diagnostics and Telemedicine of the Moscow Department of Health and the city Department of Information Technologies. Testing of complex solutions is one of the main areas of development the 2023 experiment.

Moscow has been engaged in the digitalization of the health care system for more than 10 years. The use of computer vision in medicine reduces the time spent on diagnostic procedures, and also provides medical staff with information to make more accurate diagnoses and prescribe more effective treatment. In turn, the implementation of the project on the introduction of computer vision technologies in healthcare allows you to create and develop a market for systems supporting medical decisions in radiation diagnostics.

The project meets the goals and objectives of the national project "Healthcare" and is aimed at optimizing the quality and availability of medical care for residents of Moscow.

Artificial intelligence now helps Moscow doctors diagnose multiple sclerosis

Another artificial intelligence service that recognizes multiple sclerosis on the magnetic resonance imaging brain has begun to work in Moscow. Timely diagnosis and prescription of therapy can significantly slow the progression of this chronic autoimmune disease of the nervous system. In total, in May 2023, artificial intelligence in radiation diagnostics finds pathologies in 21 clinical areas. This was announced on May 17, 2023 by the Deputy Mayor of Moscow for Social Development Anastasia Rakova.

We continue to expand the capabilities of artificial intelligence in Moscow healthcare. Now neural networks help doctors in finding signs of multiple sclerosis in radiation images. This dangerous neurological disease usually begins to develop at a young age and can lead to severe disability over time. Artificial intelligence technologies will allow doctors to increase the speed and accuracy of its diagnosis on MRI brain. Thus, smart algorithms are already assisting doctors in the search for pathologies in 21 clinical areas, - said the deputy mayor.

Since 2020, smart algorithms have been working as part of an experiment to introduce technologies computer vision in. medicine The project is being implemented on the basis Center for Diagnostics and Telemedicine of the Moscow Department of Health of. Department of Information Technology Neural networks integrated into a single radiological information service EMIAS compile a description and mark the areas of possible pathologies in the medical image with color hints, make measurements that are of clinical value to doctors when radiologist preparing a diagnostic conclusion.

Artificial intelligence helps to find in the images of radiation studies signs, lung cancer, COVID-19 osteoporosis spine, aortic aneurysm, coronary disease, hearts stroke pulmonary hypertension, hydrothorax, as well as herniated breast cancer spine, urolithiasis, arthrosis, flat-footedness and other diseases.

In Moscow, AI learned to find an ischemic stroke on CT of the brain

B To Moscow artificial intelligence learned to determine the signs of ischemic stroke on. computed tomography brain Thus neuronets , they are looking for pathologies in 20 clinical areas. Diagnostic tools based on artificial intelligence are actively developing: in less than a year, the number of directions in which algorithms work as of March 2023 has increased by more than 2 times. This was announced on March 24, 2023. NPCC DiT DZM (Scientific and Practical Center for Diagnostics and Telemedicine Technologies of the Moscow Department of Health)

For three years now, the Center for Diagnostics and Telemedicine has been testing computer vision technologies. The experiment is successful. During this time, 10 original national standards for the use of AI in healthcare, developed by our specialists, came into force, and the production of so-called phantoms, imitating organs, for training specialists and setting up medical equipment, is being launched, "said Moscow Mayor Sergei Sobyanin.

The experiment on the introduction of computer vision technologies has been carried out on the basis of the Center for Diagnostics and Telemedicine of the Moscow Department of Health together with the Department of Information Technologies for 3 years. As of March 2023, it operates on a stream of 48 artificial intelligence services from 20 development companies.

As world experience shows, algorithms that have high accuracy in diagnosing ischemia can significantly speed up the diagnosis even in the conditions of flagship hospital centers with an orientation towards emergency vascular pathology. And therefore, we really waited for the first service in this area to enter the Moscow experiment in order to assess how such a tool could be useful to Moscow healthcare. It is not always easy for radiologists to reach consensus in determining early CT signs of stroke. Automating the diagnosis of acute cerebral circulation disorders with neural networks can speed up the diagnosis process and make it more accurate, which will provide the patient with the necessary help in the shortest possible time. We now have to test this in an experiment, - said Yuri Vasiliev, chief freelance specialist in radiation and instrumental diagnostics of the Moscow Department of Health, director of the Center for Diagnostics and Telemedicine.

With the help of AI, more than 9 million radiation studies were analyzed. 44 AI services are available in 19 directions

Moscow summed up the results of three years of experiment on the introduction of computer vision in radiation diagnostics. As of mid-March 2023, 44 artificial intelligence services in 19 areas are available to radiologists in the capital. Neural networks have already analyzed more than 9 million radiation studies, Zdrav.Expert was told on March 15, 2023 at the Center for Diagnostics and Telemedicine of DZM. Such technologies accelerate patient outcomes and save physicians time describing medical images. Read more here.

Neural network learned to detect up to nine pathologies on a computed tomogram at the same time

To Moscow In expanded the possibilities of a comprehensive service - artificial intelligence now computed tomogram neuronet it can simultaneously identify up to nine pathologies. In addition to signs,, lung cancer COVID-19 osteoporosis spine, aneurysms thoracic region, aortas pulmonary hypertension, hydrothorax, coronary disease by hearts the degree of calcification of the coronary arteries and the volume of paracardial fat, the instrument also learned to detect pulmonary emphysema. This was announced on March 6, 2023, Anastasia Rakova by the Deputy Mayor of Moscow for Social Development. for almost Artificial intelligence three years now he has been helping metropolitan doctors find the slightest deviations from the norm in the pictures.

In September 2022, we launched a comprehensive service that detects several pathologies in CT scans in a matter of minutes. Let me explain by example how this works. The patient was sent to the study to exclude coronavirus pneumonia, and artificial intelligence, having analyzed the picture, found additional signs of pulmonary emphysema, which was not the purpose of the examination. This tool helps the doctor not to miss significant pathologies, - explained Anastasia Rakova.

Algorithms mark the region of possible pathologies in the medical image with color hints, make measurements that are of clinical value to the doctor when diagnosing the patient, and compile a description. An important advantage of computer vision technology is the ability to detect pathologies at an early stage, when they may be invisible to the human eye.

As part of an experiment to introduce technologies based on computer vision, neural networks work with a large flow of research, constantly receive feedback from radiologists - this is a great help for the development of services. The experiment aims to create a reliable assistant for the doctor. Over the past year, the number of doctors in Moscow who actively use algorithms in everyday practice has grown 1.5 times. I note that colleagues from the regions also use assistants. Since 2022, the doctors of the Yamalo-Nenets Autonomous Okrug, through the own development of our Center - HUB Telemed, have connected to the best artificial intelligence services in radiation diagnostics for March 2023, "said Yuri Vasiliev, chief freelance specialist in radiation and instrumental diagnostics in Moscow, director of the Center for Diagnostics and Telemedicine DZM .

Since 2020, smart algorithms have been working as part of an experiment to introduce computer vision technologies into the activities of metropolitan medical institutions. The project implements a complex of social development in Moscow on the basis of the Center for Diagnostics and Telemedicine with the support of the Department of Information Technologies.

Since 2023, in Moscow, the analysis of preventive research on mammography using neural networks has received the official status of a medical service and a special tariff. The capital has become a region in which the use of artificial intelligence algorithms is paid for from the funds of the compulsory health insurance fund.

All radiological studies in Moscow polyclinics have been transferred to digital equipment

All radiological studies in Moscow polyclinics have been transferred to digital equipment, as a result of which doctors issue conclusions in just 30 minutes instead of the previous almost 28 hours. The capital's Department of Health announced this in February 2023.

Moscow has implemented a new model for the operation of the radiation diagnostics system. Now radiologists of all adult polyclinics conduct descriptions of studies performed on digital equipment in a single round-the-clock Moscow reference center, and this is every second study in Moscow, "says Anastasia Rakova, Deputy Mayor of Moscow for Social Development. - Radiological images come from polyclinics to a single digital cloud storage of medical images, are distributed according to different modalities and specializations, and specialists immediately begin to analyze them. The location of radiologists in one building allows you to quickly conduct the necessary expert consultations in the most difficult cases.

The Department of Health said that weekly radiologists describe thousands of CT scans, MRI, fluorograms, mammograms, densitometries and X-rays at the center. In order to improve the quality of the analysis of medical images and reduce the time spent by radiologists on the description of each image, the reference center developed and implemented the concept of allocating studies to different modalities, subspecializations and internal rotation. Radiologists analyze studies within their direction, in difficult cases they can quickly get advice from colleagues. Descriptions here are made around the clock 7 days a week and do not depend on the schedule of the clinic.

By February 2023, more than 40 artificial intelligence services in 17 different areas of research are operating in the Moscow health care system. Doctors use neuronets fluorography, mammography, radiography, (computed tomography CT) and magnetic resonance imaging (MRI) for processing Artificial intelligence. helps to find on the pictures signs of lung cancer, COVID-19 osteoporosis of the spine, aneurysms aortas coronary disease, hearts stroke, pulmonary hypertension, hydrothorax, as well as breast cancer flat feet and other diseases. All^[3]

2022

Comprehensive AI service will help identify signs of coronary heart disease on CT scans

To Moscow In, they have expanded the capabilities of a comprehensive service - artificial intelligence neuronets they will detect signs of paracardial fat on CT scans, which will help in the diagnosis of cardiovascular diseases. Thus, now a smart assistant will help determine the signs of eight pathologies at once. This was announced on December 8, 2022 by the deputy head of the capital. Department of Health Ilya Tyrov

{{quote 'author
= reported Ilya Tyrov' Now a comprehensive artificial intelligence service on one CT image will measure the volume of paracardial adipose tissue. Its increase refers to independent risk factors for the development of a number of diseases, including coronary disease, hearts atrial fibrillation and metabolic syndrome, so high-quality diagnosis is very important. Thus, a smart assistant will identify signs of eight pathologies. The neural network can take over the implementation of the necessary measurements with the necessary accuracy, which facilitates the timely identification of patients from the risk group by many. to diseases Artificial intelligence in this case acts as an additional assistant and helps to reduce the burden on doctors radiologists.}}

As reported, the comprehensive service from the beginning of the launch analyzed more than 120 thousand computer tomograms of patients of medical institutions. In a series of X-rays obtained during a CT study, the neural network in a matter of minutes finds signs of lung cancer, COVID-19, spinal osteoporosis, thoracic aortic aneurysm, pulmonary hypertension, hydrothorax and coronary heart disease by the degree of calcification of the coronary arteries and the volume of paracardial fat.

As of December 2022, the comprehensive service analyzes almost every chest CT study. In just a few minutes, the radiologist receives a description from the neural network, which he can use to draw up a conclusion for the patient. The algorithm does not miss even the slightest deviations from the norm and is able to identify signs of diseases that the patient did not even suspect when referring to a doctor. As of December 2022, the development of complex services is one of the key areas of the experiment. In 2023, we plan to introduce such comprehensive neural networks in other anatomical areas, for example, to identify several pathologies at once on the abdominal CT. specified Yuri Vasiliev, director of the Moscow Center for Diagnostics and Telemedicine, chief freelance specialist in radiation and instrumental diagnostics of DMS

Neural networks are used to process images of radiation studies obtained during fluorography, mammography, radiography, computed tomography and magnetic resonance imaging. As of December 2022, the project has more than 40 services based on artificial intelligence in 17 different areas of research help radiologists in recognizing various diseases.

The algorithms work within the framework of the Experiment on the Introduction of Computer Vision Technologies into Healthcare - a project of the Social Development Complex of the Moscow City Hall on the basis of the Center for Diagnostics and Telemedicine with the support of the Department of Information Technologies.

Radiologists in Moscow began to look for pathologies of the spine and adrenal glands using artificial intelligence

Neural networks in Moscow have learned to determine the signs of two more types of diseases. Now artificial intelligence is able to diagnose protrusion and herniation of intervertebral discs on MRI, as well as an adrenal gland tumor on CT. Smart algorithms help radiologists in 17 areas of research. This was announced on November 11, 2022 by the deputy head of the Moscow City Health Department Ilya Tyrov.

{{quote 'Now artificial intelligence is able to diagnose signs of two more types of diseases - protrusion and herniation of intervertebral discs on MRI, as well as an adrenal gland tumor on CT. The algorithms work within the framework of the Experiment on the introduction of computer vision technologies into medicine, which is implemented on the basis of the Moscow Center for Diagnostics and Telemedicine. Experienced experts have already got acquainted with about 100 different solutions based on artificial intelligence, but only the best are involved in working in the city health care system. As of November 2022, about 40 services are available to Moscow doctors, which work in 17 areas of research. Not only the number of services is growing, but the list of diseases that neural networks help in diagnosing is also increasing, "said Ilya Tyrov, deputy head of the Moscow City Health Department. }}

Researchers, engineers and expert doctors test the services before allowing artificial intelligence to analyze images of radiation studies of patients in urban medical institutions. Digital assistants undergo functional and calibration tests, which check the set of declared functions, diagnostic accuracy and processing speed of one study. After the end of testing, doctors use them. As of November 2022, neural networks are available to radiologists of 150 medical institutions, including children's.

The radiologist quickly determines the signs of the disease on a radiation study, the main part of the specialist's time is taken by measurements. For example, in the presence of degenerative changes in the spine, it is necessary to estimate the percentage of stenosis, measure the size of the hernia, and so on. There may be more than 10 such measurements. The artificial intelligence service will do this quickly and accurately, and the doctor, relying on the data provided, will quickly form a conclusion, "said Yuri Vasiliev, director of the Center for Diagnostics and Telemedicine of the Moscow Department of Health.

Neural networks help to find on computer tomography signs, spine, lung cancer COVID-19 osteoporosis aneurysms of the thoracic aorta, ischemic disease, hearts stroke pulmonary hypertension, hydrothorax, and also look for breast cancer lung pathologies and many other diseases.

The experiment on the introduction of computer vision technologies into medicine was launched jointly by the Moscow Social Development Complex and the city Department of Information Technologies on the basis of the Center for Diagnostics and Telemedicine. Since the start of the Experiment, AI services have already processed more than 7 million studies. The center has become a platform for the development of artificial intelligence technologies in Russia, as well as support for domestic developers.

Thanks to the platform, all data on the state of health of citizens are available online both to doctors and to patients themselves.

A neural network for determining seven diseases at once has been introduced into the work of medical institutions in Moscow

A comprehensive service of artificial intelligence, which can immediately determine 7 pathologies, was introduced into the work of Moscow medical institutions and as of September 2022 analyzed about 28 thousand medical images. This was announced on September 8, 2022 by the deputy head of the Moscow City Health Department Ilya Tyrov.

The "forge" of artificial intelligence services is the Experiment on the introduction of computer vision technologies in healthcare, which is carried out on the basis of the Moscow Center for Diagnostics and Telemedicine. As of September 2022, 30 artificial intelligence services in various clinical areas are already operating in real time in the project, - said Ilya Tyrov.

In the spring, radiologists began testing complex artificial intelligence services that recognize seven different pathologies in one chest CT image. On a CT study, the neural network in a matter of minutes finds signs of lung cancer, COVID-19, osteoporosis of the spine, aneurysm of the thoracic aorta, coronary heart disease, pulmonary hypertension, hydrothorax.

During the experiment, due to the involvement of doctors, their feedback and the activity of the developers - participants in the experiment, the algorithms have been significantly improved. For example, the diagnostic accuracy of artificial intelligence services reaches 94%, while the proportion of false negative results is minimized. This means that neural networks do not miss even the slightest deviations from the norm. In addition, they are able to identify signs of diseases in a matter of seconds that the patient did not even know about when referring to a doctor. This means that we will improve the system of preventive examinations of citizens, which will definitely have a positive effect on their health, - said Yuri Vasiliev, director of the Center for Diagnostics and Telemedicine in Moscow.

Artificial intelligence services for automatic analysis of medical images have been tested in Moscow since 2020. As noted in the metropolitan Department of Information Technologies, after successfully checking the diagnostic accuracy and processing speed, they are connected to the Unified Radiological Information Service EMIAS, which unites all departments of radiation diagnostics of Moscow polyclinics and hospitals. Since the beginning of the Experiment on the introduction of computer vision in practical health care, neural networks have already processed more than 7 million studies.

More than a third of medical images of Muscovites are processed by neural networks

In 2022, more than 40% of radiation studies of patients Moscow in medical institutions are processed with. In artificial intelligence 2021, this figure was about 30%. This was announced on September 7, 2022 Zdrav.Expert by representatives with DMS Diagnostic and Telemedicine Center reference to the words of the deputy head. City Health Department Moscow Ilya Tyrov

"For more than two years, Moscow radiologists have had a digital assistant in their arsenal - artificial intelligence, which helps to carry out the routine work of a doctor. Neural networks analyze the radiation study in a few minutes and make the necessary measurements. Every month, about 300 thousand studies pass through smart algorithms, and in total, as of September 7, 2022, artificial intelligence services have already processed more than seven million studies. At the same time, the share of research processed by neural networks is increasing due to the expansion of the capabilities of algorithms, "said Ilya Tyrov.

Use innovative methods to analyze medical images in Moscow began in 2020. Computer vision was used to diagnose viral pneumonia and determine the degree of lung damage, then the number of directions for the operation of artificial intelligence continuously increased.

"This year, smart algorithms have learned to recognize the signs of 5 more diseases, so they help metropolitan doctors in the search for diseases in 13 areas. Artificial intelligence allows you not to miss even the slightest deviations in the medical image and becomes a doctor's assistant. Patients, in turn, receive faster and better diagnostics, "said Yuri Vasiliev, chief freelance specialist in radiation and instrumental diagnostics in Moscow, director of the Moscow Center for Diagnostics and Telemedicine.

Artificial intelligence services are integrated into a single radiological information service of a unified medical information and analytical system (ERIS EMIAS). Radiologists of all medical institutions in Moscow connected to EMIAS can use innovative technologies.

2021: Diagnostic devices of Moscow hospitals and clinics combined into a single digital network

On January 13, 2021, it became known about the unification of diagnostic devices of Moscow hospitals and clinics into a single network. As reported on the official website of the mayor of the capital of Russia, about 1300 CT and MRI devices, mammographs, densitometers, radiographic complexes, fluorographs and angiographs are connected to the Unified Radiological Information Service (ERIS).

Thanks to this, according to the Deputy Mayor of Moscow for Social Development Anastasia Rakova, during the appointment, the doctor sees the conclusion of the radiologist in the electronic medical record and, if necessary, can immediately look at the image - this helps him to diagnose or order an additional examination.

The patient receives a description of his study in the electronic medical record on the mos.ru portal and in the EMIAS.INFO mobile application immediately after the radiologist finishes filling out the protocol. Each patient's personal digital data archive is also stored in the ERIS, and doctors can contact it to track the dynamics of changes, Rakova said.

Combining diagnostic devices into a single network increases the quality of radiation diagnostics and makes it more accessible to residents of the city, explains the director of the Sergei Morozov Center for Diagnostics and Telemedicine. According to him, there is no need to go to the other end of the city if a narrow specialist is required who can perform an expert description of a difficult case: all medical images uploaded to ERIS are available to experts of the Moscow Reference Center for Radiation Diagnostics - leading radiologists, candidates and doctors of medical sciences, members of international associations.

Diagnostic equipment in hospitals in Moscow accounts for 51% of the total number of installed equipment and includes, in particular, angiographs necessary for intravascular diagnostic and therapeutic manipulations.^[4]

2020

Connecting all angiographs in Moscow to a single IT system

On December 22, 2020, it became known about the connection of all angiographic complexes (allows you to determine diseases of the cardiovascular system) in hospitals Moscow to the Unified Radiological Information Service (). ERIS We are talking about 53 systems, said the Deputy Mayor of Moscow for Social Development. Anastasia Rakova

According to her, the results of all angiographic studies are now received in a single digital archive: a medical image, description and protocol can be obtained on the day of the operation. This will allow doctors to quickly receive information about medical history, optimize operation planning, improve the quality of diagnostics, reduce intervention time and minimize risks for the patient, Rakova said, adding that all interventions can also be audited using the service.

It is noted that approximately every 20th patient who did CT or MRI angiography requires endovascular surgery. It is carried out in the departments of X-ray endovascular methods of diagnosis and treatment in Moscow hospitals under the compulsory health insurance policy. When planning operations, X-ray surgeons (vascular surgeons) and cardiac arrhythmologists need to quickly access the full history of radiation studies of patients. This is made possible by ERIS.

The basis of any decision on the operational intervention of retgenosurgeons is the visualization of the vascular bed. It is not uncommon for studies to be performed in one institution, and for an emergency operation, the patient ends up in another. Only information technologies can provide a high speed of information transfer, on which the doctor's tactics depend, - said Dmitry Skrypnik, chief specialist in X-ray endovascular diagnostics and treatment of the Moscow Department of Health.[5]

CT imaging analysis with Care Mentor AI

On October 27, 2020, Care Mentor AI announced that the AI service of the same name, analyzing CT studies and detecting signs of coronavirus infection on them, was connected to the Unified Radiological Information Service (ERIS EMIAS). The Care Mentor AI system determines the percentage and degree of lung damage in patients with COVID-19 and issues imaging of finds on CT lung series. Read more here.

Implementation of CT Image Analytics for COVID-19 technology in ERIS

On July 23, 2020, the company Crayon announced that it had joined the project for application AI health care in the city's system. Moscow

According to the company, Crayon acted as an integrator of the CT Image Analytics for COVID-19"" solution, into the ERIS (Unified Radiological Information Service) system, which is deployed by the center GBUZ "NPCC DiT DZM" and operates with the support of the department.. information technology Governments of Moscow here

Connection to ERIS of angiographs of four hospitals in Moscow

On July 9, 2020, on the official website of the mayor of Moscow, it was announced that all angiographs (for CT and MRI diagnostics of vascular pathologies) in the city's clinics will be connected to the Unified Radiological Information Service (ERIS).

By July 9, angiographs of four hospitals in the capital were connected to this IT system; by the end of September, everyone will be connected (there are 53 of them). This will optimize the planning of operations, improve the quality of diagnosis, reduce intervention time and minimize risks to the patient.

Specialists have the opportunity to consult in advance and in real time with colleagues working with the service. You can get quick access to any previously conducted study and make the necessary decision, even while in the operating room, - said Deputy Mayor of Moscow Anastasia Rakova.

He added that a single digital space for health workers and patients is being created in the capital. The large-scale project includes the introduction of a single radiological information service in hospitals. As of July 9, 2020, 1106 units of diagnostic equipment are connected to the Unified Radiological Information Service.

According to the chief specialist in radiation and instrumental diagnostics of Moscow, director of the Sergei Morozov Center for Diagnostics and Telemedicine, whose words are also quoted on the website of the mayor of Moscow, angiographic CT and MRI studies are carried out to diagnose vascular pathology. Based on the results of these studies, a decision is made on the need for surgery.

Previously, when hospitalized, the patient needed to provide the results of studies, discharges of previous hospitalizations and consultations in paper or electronic form. Now, at the stage of planning the operation, a specialist in X-ray endovascular diagnostics and treatment, a vascular surgeon have access to the archive of medical images of the patient and his medical documentation in electronic form at EMIAS, Morozov emphasized.^[6]

Launch of the X-ray analysis service using the Care Mentor AI neural network

in Artificial intelligence the capital's medical organizations has been working for the second month and now, in addition to computed tomograms the chest organs, it will also begin to analyze. roentgenograms To do this, the Unified Radiological Information Service (ERIS), to which diagnostic equipment is connected, Moscow was integrated software Care Mentor AI that analyzes and screens chest X-rays for the presence of various pathologies, including such socially significant as,, lung cancer tuberculosis pneumonia. This became Zdrav.Expert known on July 3, 2020.

Care Mentor AI is Russian a developer of services computer vision for the analysis of studies in the field of radiation diagnostics based on. neural network algorithms Its product - the service "Radiological screening of the chest organs" - is built into the working circuit of ERIS and is connected to 17 medical organizations in Moscow.

This service is designed to screen for lung pathology. It analyzes radiographic images and within a few seconds determines the presence or absence of pathological changes in the chest organs by eight groups of signs (including socially significant diseases). It also conducts medical triage, ranking research on how serious the pathology is and how quickly it needs to be looked at.

AI models are trained on a large (10k +) volume of radiological studies marked by professional radiologists. Both classification models and segmentation models were used. For each group of radiological signs of pathology, a separate neural network was trained. The predictions of these neural networks are combined using a second-level model.

The start of work with medical organizations in Moscow is an important step for our team, which allowed us to participate in practical assistance to a large number of diagnostic doctors and patients. Of course, the serious and multifaceted challenge associated with the spread of the coronavirus forced us, as an AI developer for medicine, to look at our services differently. So, we trained our neural network system to determine signs of viral pneumonia on chest X-rays, which will certainly expand its functionality and allow it to be used in mass examinations of patients during periods of seasonal outbreaks of respiratory diseases, says Pavel Reutberg, co-founder of Care Mentor AI.

The integration of services based on computer vision algorithms into the city's radiation diagnostics system takes place as part of a large-scale Moscow experiment launched by the Moscow Department of Health and the Center for Diagnostics and Telemedicine in February 2020. Its goal is to comprehensively explore the capabilities of software (software), created on the basis of computer vision technologies, when working in the conditions of city polyclinics in the capital.

The Center for Diagnostics and Telemedicine notes that before accessing ERIS, services must undergo functional testing and calibration. During functional testing on five test studies, the compliance of the AI service with the parameters declared by the developers is assessed.

The stability of each function during testing is repeatedly checked. Further, according to 11 main criteria, it is analyzed how much the AI service meets the declared parameters. Based on the results of functional testing, critical or non-critical remarks are generated. If there are no criticisms or they can be promptly eliminated, then the service can be implemented in the main, working loop of ERIS. The Care Mentor AI service successfully passed the listed stages of testing and became the first in X-ray studies, - says Sergei Morozov, Chief Specialist in Radiation and Instrumental Diagnostics of Moscow, Director of the Center for Diagnostics and Telemedicine of the Moscow Department of Health.

Care Mentor AI was one of the first to enter the "working" stage of the experiment, and another service of the company - "Mentor: CovidCT" to determine the volume and severity of lung damage in patients with COVID-19 - is in the first stage of the experiment and is ready to move to technical integration.

MR- and CT-tomographs are connected to ERIS in 64 polyclinics in Moscow

As of July 2020, ERIS combines magnetic resonance and computer tomographs in 64 polyclinics in Moscow, including outpatient institutions in Zelenograd. In total, 97 units of high-tech medical equipment are connected to the system. Digital devices of classical X-ray and devices for fluorography are connected to the system.

The world's largest base of CT examinations of patients with coronavirus has been collected in Moscow

On May 26, 2020, it became known that the world's largest base of CT studies of patients with coronavirus COVID-19 was collected in Moscow. It included more than a thousand completely impersonal sets of computer tomograms registered from March 10 to April 25 as part of a single radiological information service, said Anastasia Rakova, deputy of the capital for social development.

To form a database of medical photographs, on the basis of which artificial intelligence can be trained, about 80 thousand studies were analyzed, which were made to patients with suspected viral pneumonia in Moscow outpatient CT centers (by May 26, 2020 there are 48 of them).

Dataset is available for artificial intelligence developers. Computer tomograms have been converted into the NIfTI format recognized among machine learning professionals.

The dataset, collected by experts at the Centre for Diagnostics and Telemedicine, is said to be unique. For example, the UC San Diego database contains 349 2D CT scans of 216 patients, while the datacet collected in Moscow contains 3D CT scans. The RAIOSS & Livon Sa de database, formed in Brazil, so far includes 10 CT scans, while the Italian Radiological Society has more than 70. The set of cases of the new coronavirus infection of the North American Radiological Society is scattered and suitable only for familiarization, and the British Society of Thoracic Radiology contains no more than a hundred studies.

In addition, it is noted that all CT studies in the datacet of the Center for Diagnostics and Telemedicine are marked according to the classification reflecting the volume of pathological changes in lung tissue in coronavirus according to computed tomography of the chest organs.^[7]

All polyclinics and hospitals in Moscow are connected to an AI service for diagnosing COVID-19 by CT scans

On May 12, 2020, it became known that all polyclinics and hospitals in Moscow are connected to an AI service for diagnosing COVID-19 from CT images. In two weeks, about 30 thousand examinations were processed through this system, said Anastasia Rakova, Deputy Mayor of Moscow for Social Development, whose words are quoted on the official website of the capital's mayor.

During an epidemic in the city, technologies for automatic data processing and artificial intelligence help save the resources of medical personnel and speed up the time for diagnosing patients... The technology is actively used in hospitals and outpatient CT centers, where the main stream of patients with suspected COVID-19 is now concentrated, she said, adding that the accuracy of determining the stages of pneumonia development from CT scans exceeds 90%.

By downloading a CT image on their work computer, doctors see a preliminary conclusion made using artificial intelligence technologies. If there are signs of pneumonia, the result of the study is highlighted in red, and areas of possible damage are marked directly on the image of the lungs. The final diagnosis is made by the doctor.

Artificial intelligence programs work on the basis of the Unified Radiological Information Service (ERIS), to which all clinics and hospitals are connected. Doctors of private and federal COVID centers have access to them.

By May 12, 2020, 48 outpatient CT centers are operating on the basis of the buildings of city polyclinics in Moscow to diagnose patients with ARVI, community-acquired pneumonia and suspected new coronavirus infection. As of May 12, 121,301 cases of coronavirus COVID-19 and 1,179 deaths from this infection were recorded in the capital.^[8]

Allocation of 147.3 million rubles for the introduction of AI in ERIS

In February 2020, the Moscow Department of Health announced the allocation of 147.3 million rubles for the introduction of artificial intelligence technologies into the capital's unified radiological information service (ERIS).

In total, the government approved 16 grants to medical software developers to test decision support methods based on data analysis results in three areas: computed tomography and low-dose chest computed tomography (CT/NDCT); mammography (MMH); radiography (RG). 

The experiment will be carried out in the mode of interaction with the Department information technology Moscow (DIT) and involves, among other things, the introduction of additional settings and improvements to the cloud Unified Radiological Information Service (ERIS) as part of an automated one. GIS "Unified Medical Information and Analytical System of Moscow"

Developers of ready-made solutions are allowed to participate in the pilot project, which will last until the end of 2020, who have documents confirming the compliance of the service quality with FDA, CE or their analogues, reports on clinical testing or application of the solution to medical organizations, as well as scientific articles on assessing the accuracy and effectiveness of the service.

After admitting the service to participate in the experiment, grantees will have to upload the results of radiological studies to the ERIS so that they can be remotely described by a group of experts.

The data obtained during the experiment will be considered by a special commission at the Scientific and Practical Clinical Center for Diagnostics and Telemedicine Technologies of DZM. Based on the results of the study of the results of the experiment, the commission will decide on the introduction of the selected technologies.^[9]

ERIS Capabilities and Objectives

An important area in the development of ERIS as of 2020 is the integration of the system with EMIAS. Integration will allow combining patient data, clinician prescriptions, and research findings. This, in turn, will increase the quality of solving diagnostic problems by radiologists, facilitate the work of clinicians, and improve patient care.

The main tasks of ERIS are to increase the efficiency of radiation diagnostics in Moscow, create a single network that unites diagnostic equipment, and provide a modern system for storing images, descriptions and conclusions obtained as a result of research.

The ERIS system allows:

• conduct analysis of the quality of performance and interpretation of studies through control - regular audits;
• improve the quality of diagnostics - due to the professional growth of radiologists and consultations of remote experts - the best specialists in radiation diagnostics;
• improve the qualifications of specialists - through research to outline a training program for nurses and doctors;
• introduce a unified standard for the quality of radiation diagnostics in accordance with the leading global trends in all hospitals and clinics of the DSM;
• make a request for advice in complex clinical cases.

For the city, ERIS makes it possible to analyze in real time the effectiveness of the medical equipment used in conducting radiological studies, redirecting patients to the least busy devices.

An additional feature of ERIS is the transparency and completeness of diagnostic information. Thanks to access to all created descriptions and conclusions, SPC experts medical radiology can conduct selective audits, analyzing the created protocols, identifying typical errors and shortcomings, correcting clinically significant discrepancies.^[10]

2018: Protection of personal data of patients of radiological offices in polyclinics in Moscow

On September 17, 2019, Elefus announced that it had developed and implemented an IT system for protecting personal data that is processed and stored in a single radiological information service (ERIS). Work on the project was completed in 2018. Read more here.

Notes