April 22 2022
Reading time 12 minutes
Artificial intelligence (AI) solves problems that require human thinking abilities in information processing and remembering. Due to machine learning and computer vision methods, technologies analyze big volume of data within short periods identifying mathematical patterns in them.
In 2019, the EY company has conducted a research of commercial medicine market and healthcare system in Russia. A survey has shown a growing interest to healthcare digitalization: 38% of respondents notes that technological effectiveness of medical assistance is an indispensable criterion for its quality1.
At the same time, physicians face the necessity to process a large volume of information, such as patient’s personal data, history and test results. Examination is also usually associated with many routine tasks, for instance, in medical visualization when interpreting diagnostic images.
The COVID-19 pandemic has highlighted the need to find new approaches to medical care. In this period, regional spending on healthcare increased almost doubled, RBK reports with reference to the Analytical Credit Rating Agency (ACRA)2.
Artificial intelligence-driven solutions are seamlessly built into the creation of a digital health circuit:
prompt information processing and transmission;
partial automation of testing;
an opportunity to compensate for the shortage of qualified employees.
In July 2020, a pilot project was completed in Kirov Region to introduce a clinical decision support system. Medical Information and Analysis Center and a number of regional medicine organizations, including Kirov Clinical Diagnostic Center, the Center for Oncology and Medical Radiology, and the Center for Cardiology and Neurology, participated in the project implementation3.
Based on its results, the participants noted the importance of the artificial intelligence prediction tools3:
the system assesses risk factors of a disease;
it helps to detect pathology in its early stages;
it simplifies preventive measures for chronic non-communicable diseases.
Further integration of AI technologies in medical information systems is underway. It is planned that a patient will be able to see the results of examinations in the personal account3.
In the Republic of Sakha a pilot project ‘ONKOPOISKSAKHA.RF’ (onco-search Sakha) was launched with the support from the Ministry of Health of Russia. The developers are considering the use of AI solutions in preventive medicine for oncopathologies. The official website of the online project shows the main stages of screening4:
a patient fills in a questionnaire and a neural network assesses it;
If there is a high oncology risk, the system suggests making an appointment at the clinic;
the patient receives basic laboratory and instrumental examinations at the Yakutsk Republican Oncologic Dispensary;
based on the questionnaire and clinic visit, an additional examination can be prescribed.
The project is aimed at the identification of widespread oncological diseases4:
Single Radiology Information System (SRIS) of Moscow provide for the efficient exchange of information between diagnostic centers. Its database contains data from radiological examinations: CT, MRI, photofluorography and radiography5.
Various AI platforms with computer vision technology are integrated into the service. The algorithms provide a “second opinion” to the radiologist accelerating the diagnostic image annotation. The AI identifies pathological areas, assesses the extent of the damage, and generates a preliminary description6.
In July 2020, the Ministry of Health of the Russian Federation supported the development of standards in the application of artificial intelligence in health care. The agency pointed out that the creation of a regulatory framework will help to develop common approaches to safety assessment and reliability of AI-based systems7. The initiative is aimed at increasing trust in the use of technologies in medicine8.
In Autumn 2020, the Digital Diagnosis project was launched with the support of Ministry of Health of Russia, according to the agency. Thirteen medical organizations of the Khanty-Mansi Autonomous Territory participated in the experiment9.
A number of advantages of the AI use were noted by the participants during the project, for instance:
it draws the attention of the physician to pathological changes;
it helps to formulate the final opinion;
it accelerates the examination by 4-5 times;
it compensates the shortage of qualified experts.
Experts have studied over 2,000 diagnostic images using the computer vision technology9.
The Ministry of Health, together with Rostec Corporation, has launched a project to create a platform that provides information support to the AI solutions developers in the field of medicine. The result was the first version of the federal AI platform, created in July 2021, reports Kommersant with reference to the agency10.
Training and testing the algorithm require verified datasets. The developers were provided with the information integrated from electronic health records. All data from Russians was anonymized10.
The sources of the data included10:
Data from the Federal Integrated Electronic Health Record, which is a subsystem of the Single State Information System in Healthcare.
Information from medical research centers under the Ministry of Health: National Medical Research Center of Neurosurgery named after Academician N. N. Burdenko, National Medical Research Center of Surgery named after A. V. Vishnevskiy, National Medical Research Center of Oncology named after N. N. Blokhin, National Medical Research Center named after V. A. Almazov and National Medical Research Center of Endocrinology.
The collected information helped to form annotated datasets, including conclusions of radiological diagnosis and pathomorphological studies10.
Relevant areas of application of artificial intelligence in medicine are implemented by SberMedAI. The eco-system integrator offers an AI-based Medical Digital Diagnostic Center (MDDC) to health workers and health care organizers. MDDC includes algorithms that solve a wide range of clinical tasks:
TOP 3 smart assistant for doctors. The service analyzes data from patient’s electronic health record and history. The AI offers three most likely diagnoses to the physician using alphanumeric code in accordance with ICD-10.
CT Stroke. The algorithm marks zones of impaired blood circulation, which may occur with stroke, on the head CT scans. Changes are assessed using the ASPECTS scale.
CT of Lungs. The artificial intelligence lights the pulmonary tissue foci, which emerge with respiratory diseases, including COVID-19 viral pneumonia. Another task of the algorithm is early oncological diseases diagnosis. The neural network identifies minimal nodules exceeding the limit of 4 mm.
Mammography. The service raises screening efficiency: the algorithm highlights the contours of neoplasms and assesses changes on the BI-RADS scale. The physician sees the probability of malignant tumor calculated by the system. This helps to plan further examination and treatment.
Short ECG. The AI accelerates the provision of assistance to patients in grave condition: ambulance and urgent care center physicians analyze ECG faster in the form of online consultation.
AI Resp mobile app. The artificial intelligence processes the uploaded audio recording of breath and speech and identifies acoustic signs of respiratory infections.
AI Skin mobile app. A patient can make an initial diagnosis of skin diseases on their own. They only need to take a photo of a problem area and upload it. The service assesses the probability of 7 skin rashes, and the conclusion is checked out during a visit to a physician.
Geography of successful MDDC algorithms implementations in Russian regions:
The Karachay-Cherkess Republic. A pilot project CT of Lungs Examinations Retrospective Analysis was carried out. Earlier, SberMedAI implemented the CT of Lungs algorithm to detect COVID-19 in the Karachay-Cherkess Republic. 1,500 anonymized images have been analyzed for the second time: now the AI was searching for the nodules in lungs. As a result, pathological neoplasms were detected in 12 patients, among whom 8 were confirmed by oncologists as cancerous.
Yaroslavl Region. 55 ambulance station and urgent care center were equipped with the MDDC Cardio service. The cardiogram analysis by the artificial intelligence accelerated the ECG decoding in the village of Porechye-Rybnoye.
Tyumen Region From November to January 2022 over 500 CT examination of brain was carried out in the Regional Teaching Hospital No.2 using the CT Stroke service.
Khabarovsk. CT Stroke has significantly reduced the time it takes to make a diagnosis. It used to take from 15 to 60 minutes for a physician of the Territorial Teaching Hospital No.2 of Khabarovsk to describe a brain scan. Using the AI capabilities, specialists of the Territorial Teaching Hospital No.2 have raised the speed of diagnosis: now a preliminary diagnosis can be made in 2 minutes.
Nizhniy Novgorod Region. SberMedAI helped to introduce new solutions in the Central District Hospital of Bor. Computed tomography scanners have been equipped with the CT of Lungs software. During the piloting of the project, more than 100 studies were conducted. Patients’ routing was implemented due to the service: The AI helps to faster accelerate identification of patients in grave condition.
The artificial intelligence use facilitates the improvement of treatment and diagnostic process. The attention of Russian AI developers is directed at various spheres of diagnosis, from screening to in-depth symptom analysis with a verified conclusion.
A showcase of endoscopic examination digitalization is the implementation of artificial intelligence in the City Teaching Hospital named after Botkin. The service testing was launched in February 202111.
The used AI is trained on endoscopic images received from practice during photo or video recording. The experience of numerous researches is transformed into digital algorithms, which are already helping to solve diagnostic tasks11.
The service facilitates visual search for pathologies: The AI lights pathological area, drawing physician’s attention. The algorithm provide qualified support to young physicians, raising diagnostic visual experience11.
In November 2021, the Intellogic company launched a pilot project in Alexandrovskaya Hospital in St. Petersburg using digital medical services, RBK reports12.
Botkin.AI platform facilitates efficient responses to ‘COVID-19-related’ calls and saves human and material resources12:
it analyzes history data and data from laboratory and instrumental examinations of patients;
it creates a mathematical model showing individual pathogenetic chains;
it allows physicians to select the therapy and to assess the likely results of treatment;
it makes a forecast about the probability of death after a surgery or other intervention.
The service minimizes medical errors in cancer screening, helping to detect pathology at an early stage12.
Emergency Medical Care Station A.S. Puchkov in Moscow uses artificial intelligence for prompt assessment of calls from patients13.
The algorithm of medical and information complex work is as follows13:
The dispatcher accepts a call and fills in a questionnaire that notes the patient’s answers about their complaints.
The artificial intelligence assesses the data from the questionnaire and assigns a reason: emergency or immediate.
It reduces the time for assigning an ambulance or urgent care center team.
The development has reduced the time of call processing and has received a patent, Healthcare Department of the city of Moscow13.
The further development of intellectual systems in Russia can be associated with the following steps:
The creation of the AI use legal basis. The development of regulations on the use of artificial intelligence will eliminate possible legal ambiguity.
Efficiency assessment and report publication. The objective data on diagnostic capabilities of methods and their application limits will enhance confidence among both physicians and patients.
AI startups support. Attracting investments and increasing public funding play an important part in IT sector development. A simplified registration system for medical devices and provision of consulting made it easier for developers to enter the market.
Medical community engagement. It is impossible to tell physicians about advances in healthcare technology and opportunities for professional growth.
The active interaction with research centers and medical institutions. Training and testing algorithms require big volumes of verified data. Hospitals can provide developers with anonymized histories, laboratory examination results and descriptions of diagnostic scans. Expert physicians can function as opponents to the AI and provide feedback on service improvement.
Prospects of the AI in healthcare are prompt and accurate examinations and higher quality of medical assistance. An important condition of development is an active dialogue between all parties engaged, namely, developers, physicians and patients.
The Overview of Commercial Medicine Market of Russia in 2018-2019 [Online]. URL: https://assets.ey.com/content/dam/ey-sites/ey-com/ru_ru/news/2020/03/ey_healthcare_research_2018-2019_24032020.pdf (accessed 02/03/2022).
Healthcare Involves Artificial Intelligence: RBK [Online]. URL: https://plus.rbc.ru/news/60b769367a8aa93e70361a37 (accessed 02/03/2022).
Artificial Intelligence Used in Medicine in Kirov Region: Ministry of Health of the Russian Federation [Online]. URL: https://minzdrav.gov.ru/regional_news/14504-iskusstvennyy-intellekt-primenyaetsya-v-meditsine-kirovskoy-oblasti (accessed 02/03/2022).
Onco-search: ONKOPOISKSAKHA.RF (Onco-search Sakha) [Online]. URL: https://xn--80aawhbpgcajyb7c.xn--p1ai/about (accessed 02/03/2022).
Single Radiological Information System Created in Moscow: Official Website of the Mayor and Government of Moscow [Online]. URL: https://www.mos.ru/mayor/themes/20299/3374050/ (accessed 02/03/2022).
Over 150,000 Radiological Scans Processed by Artificial Intelligence in Moscow: Healthcare Department of the City of Moscow [Online]. URL: https://mosgorzdrav.ru/ru-RU/news/default/card/4402.html (accessed 02/03/2022).
Federal Agency on Technical Regulating and Metrology [Online]. URL: https://www.rst.gov.ru/portal/gost/home/presscenter/news?portal:componentId=88beae40-0e16-414c-b176-d0ab5de82e16&navigationalstate=JBPNS_rO0ABXczAAZhY3Rpb24AAAABAA5zaW5nbGVOZXdzVmlldwACaWQAAAABAAQ3MTA0AAdfX0VPRl9f (accessed 02/03/2022).
Ministry of Health of Russian Supports Development of Main Standards for Artificial Intelligence Systems in Clinical Medicine: Ministry of Health of the Russian Federation [Online]. URL: https://minzdrav.gov.ru/news/2020/07/28/14564-minzdrav-rossii-podderzhivaet-razrabotku-osnovopolagayuschih-standartov-po-sistemam-iskusstvennogo-intellekta-v-klinicheskoy-meditsine (accessed 02/03/2022).
Yugra Experience of Artificial Intelligence System Introduction in Healthcare Recognized as One of the Most Cutting-Edge in the Country: Ministry of Health of the Russian Federation [Online]. URL: https://minzdrav.gov.ru/regional_news/16931-opyt-yugry-po-vnedreniyu-sistem-iskusstvennogo-intellekta-v-zdravoohranenii-priznan-odnim-iz-peredovyh-v-strane (accessed 02/03/2022).
Artificial Intelligence to Take Care of Health: Kommersant [Online]. URL: https://www.kommersant.ru/doc/4907745 (accessed 02/03/2022).
Hospital Named After Botkin Started Testing Artificial Intelligence to Analyze Endoscopic Examination: City Teaching Hospital named after S. P. Botkin, Healthcare Department of the City of Moscow [Online]. URL: https://botkinmoscow.ru/news/v-botkinskoj-nachali-testirovat-iskusstvennyj-intellekt-dlya-analiza-endoskopicheskix-issledovanij/ (accessed 02/03/2022).
Project to Combat Medical Errors Launched in St. Petersburg: RBK [Online]. URL: https://www.rbc.ru/spb_sz/22/11/2021/619b7daa9a7947aa2c03b777 (accessed 02/03/2022).
Moscow Ambulance Developed Unique Patient Inquiry Algorithm: Healthcare Department of the City of Moscow [Online]. URL: https://mosgorzdrav.ru/ru-RU/news/default/card/4823.html (accessed 02/03/2022).
Computed tomography (CT) facilitates early diagnosis of a wide range of diseases. When carrying out CT,...
April 29, 2022
Software as a medical product Use of software, including AI-based software, can impact clinical decision making....
April 29, 2022