Transition to electronic medical records (EMR) is one of the basic directions of digital transformation of healthcare.

One of the urgent modern problems of EMR management is the quality of data that are accumulated in modern medical information systems. Given the growing role of EMRs as a source of information for medical decision support systems, the introduction of management elements based on primary data, and the development of research in the field of real-world clinical practice data (RWD), there is a growing need for reliable and objective methods to assess the quality of data accumulated in EMRs. In this regard, the development of reliable methods and tools for data quality assessment (DQA) in EMR is an urgent scientific task.

Aim. To study and systematize the approaches, methods and criteria of proposed in the scientific literature.

Materials and Methods. Reviews and original articles on the subject of EMRs DQA were studied. Sources were identified by systematic search in four electronic bibliographic databases: PubMed, Web of Science, Scopus and RSCI.

Results. The paper presents the main approaches and criteria for assessing the quality of EMR data, harmonizes the terms and definitions of DQA, and identifies the key components required to implement an EMRs DQA system.

Conclusion. The generic EMRs DQA criteria formulated in the review can be used for further research and development of DQA tools, including by medical information system developers and health care organizers responsible for the digital transformation of the industry. Also, this work will help eliminate confusion about EMR data quality management and provide the guidance needed to develop effective DQA programs.

Aim. The purpose of this article is review and analyze methods for measuring cognitive load, as well as approaches to using machine learning techniques to identify EEG data.

Materials and methods. The review systematizes and summarizes the information on the topic under consideration. Scientific articles were searched in bibliographic databases: eLIBRARY, ScienceDirect, Scopus.

Results. This review focused on ways to measure the cognitive load of the brain, modern EEG recording devices, and methods for transforming, extracting, and classifying features from acquired EEG signals.

Conclusion. With new wearable devices available for acquiring and processing EEG signals, there is a need to develop new approaches for using machine learning to identify cognitive brain processes.

Introduction. The introduction of artificial intelligence (AI) technologies in Russian healthcare is an important step to improve the efficiency and quality of medical care. The development of AI technologies helps automate data processing, support physician decision-making and improve predictive analytics.

Aim. Analysis of the results of creation and implementation of software solutions using AI technologies in Russian healthcare. Materials and Methods: A systematic search for data in the State Register of Medical Devices and on the official website of the Unified Information System in the field of procurement was conducted. The main research methods were analysis of registered AI-based medical devices and monitoring of their use in medical institutions.

Results. Quantitative indicators of implementation of solutions using AI technologies in Russian healthcare have been determined. The factors facilitating and hindering the introduction of innovations were formulated. The list of components of the methodology of implementation and operation of medical solutions based on AI technologies, related changes in the organization of medical care, personnel training, and patients' involvement in the development of their health has been defined.

Conclusions. Significant progress in the use of AI in Russian healthcare requires further disclosure of methodological, organizational, technological and economic issues. Continued sharing of regional practices and knowledge will be key to building trust in AI technologies.

State programs in the field of healthcare informatization implemented in 2011–2021 resulted in over 91% of state and municipal medical organizations implementing various medical information systems. This made it possible to start the transition to electronic medical records (EMR). The extraction of real-world data (RWD) from the accumulated EMRs and subsequent analysis of these data opens new and promising opportunities for the development of domestic healthcare.

Aim: to analyze the RWD extracted from anonymized EMRs.

Materials and methods: we used Webiomed's predictive analytics platform database, which at the time of the study had accumulated anonymized EMRs of over 29 million patients, including 229 million different medical documents. Data providers for the platform were 856 medical organizations from 28 regions of the Russian Federation. The functionality of the Webiomed platform allows processing unstructured medical documents, extracting data from them suitable for analysis using artificial intelligence technologies.

Results. This paper presents: analysis of medical organizations as data providers for Webiomed platform, analysis of EMC structure and composition, analysis of patient population. The analysis was performed at the moment of data upload on 16.10.2023. More than 4 billion 558 million structured attributes were extracted from the accumulated anonymized EMRs and systematized by different types. The platform contains 147,886,190 cases of diseases classified according to the ICD-10. 8,393,403 patients (28.71% of the total) have medical information in the EMR. The proportion of “empty” EMCs (which did not contain any medical record) was 71.29%. However, EMRs of 3,448,797 patients had more than 10 medical records. The structure of EMRs is dominated by protocols of medical examinations, protocols of laboratory tests, electronic prescriptions, and instrumental examinations. 4,456,263 patients have a depth of data collection of more than 3 years.

Conclusion. The results show that the extraction and processing of RWD from anonymized EMRs do allow the creation of large sets of structured data. Currently, to the best of our knowledge, the Webiomed platform contains the largest database of RWD extracted from EMRs in Russia. The material presented in this paper is the first analysis of EMRs and extracted features performed and published in Russia. Ensuring the quality of work with RWD at all stages, from the development of EMR structure and data input to the formation of digital twins, is the most important condition for their application to solve various tasks in the healthcare system and pharmaceutical industry.

Objective. The study is devoted to modern forms of case method implementation as a key tool in the development of clinical thinking of doctors. The main factors complicating the creation of situational tasks and limiting the large-scale application of this method in medical education are identified.

Materials and methods. The concept of using large language models (LLM) to reduce the complexity and labor intensity of developing situational tasks in medical education is proposed.

Results. A prototype of an interactive LLM ChatGPT-4o-based case study based on a clinical guideline for chronic heart failure has been developed and tested. The prototype allows dialogic interaction with learners, generation of laboratory and instrumental data, and real-time adaptation of case complexity. Despite its effectiveness, risks associated with the occurrence of content generation errors (so-called “hallucinations”) have been confirmed.

Conclusion. The concept of LLM application for automation and improvement of case method in medical education is proposed. Requirements for the development of a digital solution are formulated, which will greatly simplify the creation and modification of case problems and will ensure the development of clinical thinking of physicians. Further efforts should be aimed at minimizing generative errors and creating specialized interfaces for effective use of LLM in training.

Background. Diagnosis of early-stage chronic kidney disease (CKD) is a global challenge, as late-stage disease is more commonly diagnosed. The development of modeling methods for making management decisions aimed at improving the efficiency of early diagnosis of CKD is an important scientific and practical task, which can be greatly supported by the use of machine learning algorithms (MLA).

Aim. To improve the accuracy of diagnosis of CKD using data from history, clinical-instrumental, genetic examination and machine learning algorithms (MLA).

Methods. Data were obtained from a single-center retrospective catamnestic cohort study (2011–2022) of children with CKD stage 1–4 aged 1 to 17 years. The main group included 128 children with CKD, and the comparison group included 30 children without any kidney disease. Two groups were comparable by sex and age. The data of anamnesis, clinical-instrumental and genetic examination were used to build a model for CKD diagnosis. The model was built using the MLA multivariate logistic regression (MLR). Three variables were used in the model: erythrocyte sedimentation rate in blood (β = 0,392; p<0,001).

Results. A diagnostics model was obtained allowing prediction of CKD on a test sample with accuracy of 90,3% [80,6; 96,8], sensitivity of 92,0% [81,5; 100,0], specificity of 83,3% [50,0; 100,0], ROC-AUC = 90,0% [77,2; 100,0]. The resulting model is of excellent quality (>90%) as the ROC-AUC is 0,90 on the test sample. The cut-off point value of the probability of CKD is 0,25.

Conclusions. We developed and tested the model that diagnoses early-stage CKD in children with high accuracy.

Background. Diagnosis of early-stage chronic kidney disease (CKD) is a global challenge, as late-stage disease is more commonly diagnosed. The development of modeling methods for making management decisions aimed at improving the efficiency of early diagnosis of CKD is an important scientific and practical task, which can be greatly supported by the use of machine learning algorithms (MLA).

Aim. To improve the accuracy of diagnosis of CKD using data from history, clinical-instrumental, genetic examination and machine learning algorithms (MLA).

Methods. Data were obtained from a single-center retrospective catamnestic cohort study (2011–2022) of children with CKD stage 1–4 aged 1 to 17 years. The main group included 128 children with CKD, and the comparison group included 30 children without any kidney disease. Two groups were comparable by sex and age. The data of anamnesis, clinical-instrumental and genetic examination were used to build a model for CKD diagnosis. The model was built using the MLA multivariate logistic regression (MLR). Three variables were used in the model: erythrocyte sedimentation rate in blood (β = 0,392; p<0,001).

Results. A diagnostics model was obtained allowing prediction of CKD on a test sample with accuracy of 90,3% [80,6; 96,8], sensitivity of 92,0% [81,5; 100,0], specificity of 83,3% [50,0; 100,0], ROC-AUC = 90,0% [77,2; 100,0]. The resulting model is of excellent quality (>90%) as the ROC-AUC is 0,90 on the test sample. The cut-off point value of the probability of CKD is 0,25.

Conclusions. We developed and tested the model that diagnoses early-stage CKD in children with high accuracy.