Olesya Anatolievna Rubanenko

The objective was to analyze the literature sources on monitoring and observation of patients with diseases of the gastrointestinal tract (GIT) in daily medical practice using machine learning methods. Methods and materials. To prepare the review, scientific publications were searched in databases such as PubMed, Web of Science, Scopus, CyberLeninka, eLibrary, and Google Scholar. The search strategy included the use of keywords in Russian and English: «diseases of the gastrointestinal tract», «gastroenterological diseases», «artificial intelligence», «machine learning», «deep learning», «patient monitoring», «remote monitoring». The inclusion of original research in the period 2015–2025 is based on an independent assessment by the authors. Results. Of the 594 publications, 9 studies meeting the inclusion criteria were included in the final analysis after screening. Conclusion. AI provides modern approaches to monitoring, diagnosing, and predicting complications of gastrointestinal diseases. The solutions created on its basis are characterized by high diagnostic and forecasting accuracy, often exceeding classical clinical scales, and form the foundation of intelligent decision support systems for doctors.

The objective was to analyze the literature sources on monitoring and observation of patients with diseases of the gastrointestinal tract (GIT) in daily medical practice using machine learning methods. Methods and materials. To prepare the review, scientific publications were searched in databases such as PubMed, Web of Science, Scopus, CyberLeninka, eLibrary, and Google Scholar. The search strategy included the use of keywords in Russian and English: «diseases of the gastrointestinal tract», «gastroenterological diseases», «artificial intelligence», «machine learning», «deep learning», «patient monitoring», «remote monitoring». The inclusion of original research in the period 2015–2025 is based on an independent assessment by the authors. Results. Of the 594 publications, 9 studies meeting the inclusion criteria were included in the final analysis after screening. Conclusion. AI provides modern approaches to monitoring, diagnosing, and predicting complications of gastrointestinal diseases. The solutions created on its basis are characterized by high diagnostic and forecasting accuracy, often exceeding classical clinical scales, and form the foundation of intelligent decision support systems for doctors.

Aim. To analyze the phenotypes of patients at highest risk for albuminuria (AU) in a population of patients without diabetes and to estimate its prevalence in the identified phenotypes. Material and methods. A total of 4580 patients over 40 years of age without type 1 or 2 diabetes, who had not previously been diagnosed with chronic kidney disease, were included in a real-world practice registry of AU detection in patients with previously undiagnosed CKD — AURA. The registry design did not include patient follow-up. AU was determined using sensitive micral-test strips, with an AU cutoff of 20 mg/L. Results. AU ≥20 mg/L was present in 64,9% of cases. According to univariate analysis, patients with AU compared with those without it were more likely to have the following risk factors: hypertension (79% vs 71,5%), male sex (44,2% vs 38,6%), smoking (31% vs 25,2%), metabolic syndrome (32% vs 23,4%), as well as diseases such as coronary artery disease (35,5% vs 29,7%), heart failure (41,5% vs 35%), atrial fibrillation (17,3% vs 12,5%), prediabetes (14,7% vs 8,9%), chronic obstructive pulmonary disease (COPD) (7,67% vs 3,76%), chronic cystitis (5,35% vs 3,49%). A multivariate model revealed that age (with each 1-year increase in age, the risk of AU increased by 2%, p-value Conclusion. A multivariate analysis of AURA registry data confirmed that AU is independently associated with conditions such as post-infarction cardiosclerosis, atrial fibrillation, prediabetes, metabolic syndrome, chronic cystitis, and COPD. Of particular interest is the study and follow-up of patients with COPD, in whom AU was found in 78,9%.

Aim.  To identify the optimal combination of cardiovascular toxicity (CVT) markers for subsequent optimization of the diagnosis of cardiovascular adverse events in patients with non-Hodgkin lymphomas. Material and methods.  This prospective study was conducted to monitor cardiovascular adverse events in patients treated with the R-CHOP regimen. A total of 116 patients were divided into main (n=58, mean age 55,4±12,4 years, 25 (43,1%) men) and control groups (n=58, mean age 54,9±11,4 years, 32 (55,1%) men). Division of patients into subgroups was carried out taking into account CVT: A1 and B1 — with CVT: A1 — 21 patients, 55,9±11,17 years, 15 men (71,4%); B1 — 10 patients, 56,75±9,11 years, 6 men (60%); A2 — 37 patients, 55,16±13,12 years, 13 men (35,1%); B2 — 48 patients, mean age 54,24±12,18 years, men n=28 (58,3%). Results.  CVT was more common in men (p=0,039, p _A1-A2 =0,008, p _A2-B2 =0,031) and smokers (p=0,026, p _A1-A2 =0,04, p _A2-B2 =0,046, p _A2-B1 =0,004). The level of N-terminal pro-brain natriuretic peptide (NT-proBNP) significantly increased from 86,50 (68,50;113,50) to 150,50 (71,00;372,50) pg/ml (p=0,023, p _V2-V3 =0,026) in the target group of patients. Among patients with CVT, a longer QTc was recorded, as well as a lower level of global longitudinal systolic strain (p=0,008, Visit 2, p=0,003, Visit 3). Conclusion.  The mathematical model developed by the authors, based on paraclinical parameters and allowing for CVT diagnosis optimization, is proposed for use in real-world practice.

Aim.  To study the role of imaging technologies in reclassifying cardiovascular risk (CVR) and choosing management strategies for asymptomatic patients. Material and methods.  A total of 137 asymptomatic patients aged ≥40 years (men, 85,4%; mean age, 52,0±7,2 years) were examined. The Systematic Coronary Risk Evaluation 2 (SCORE2) was used to determine the risk of cardiovascular disease. All patients were assessed for coronary calcium score (CCS), the condition of the carotid (extracranial artery Doppler ultrasound) and coronary (coronary computed tomography angiography (CCTA)) arteries. In addition, glomerular filtration rate were estimated with subsequent CVR restratification. Results.  Depending on the CCS, patients were divided into following groups: group 1 — 78 (56,9%) patients with CCS of 0 (median age, 50,0 (44,0; 53,0) years; men, 83,3%); group 2 included 59 patients (43,1%) with CCS >0, median, 87,0 (16,5; 187,0); median age, 53,0 (49,0; 59,0) years; men, 88,1%. CCS determination, extracranial artery Doppler ultrasound, CCTA allowed for risk reclassification in 28 patients (20,4%): 10 patients (12,8%) in group 1 and 18 patients (30,5%) with a CCS >0 in group 2. In 2,9% of patients, the estimated glomerular filtration rate increased the CVR class. Risk reclassification based on imaging led to pharmacotherapy adjustments. Conclusion.  Supplementing the traditional SCORE2 with imaging techniques (CCS, CCTA, extracranial artery Doppler ultrasound) improves the CVR class in asymptomatic patients, enabling personalized primary prevention of cardiovascular events.

Aim . To review the literature on the effectiveness of artificial intelligence (AI) technologies for screening, diagnosis, and monitoring of cardiovascular diseases. Material and methods. We searched publications in PubMed, Web of Science, Scopus, CyberLeninka, eLibrary, and Google Scholar. The search strategy included the following keywords: "circulatory system diseases", "cardiovascular diseases", "artificial intelligence", "machine learning", "deep learning", "patient monitoring", "remote monitoring". Inclusion of original studies from 2015 to 2025 was based on independent author assessment. R e sults . Of 594 publications, 8 studies meeting the inclusion criteria were included in the final analysis. Conclusion . AI is a tool that is transforming modern methods of monitoring, diagnosing, and predicting cardiovascular outcomes. AI-based solutions demonstrate high diagnostic and prognostic efficacy, often exceeding traditional clinical scores, and form the basis of intelligent decision support systems for physicians.

Aim . To evaluate the literature on digital auscultation, comparing the effectiveness of human auditory system and machine analysis. Material and methods. A search of publications was conducted in PubMed, eLibrary, and Google Scholar using the following keywords: "digital auscultation", "digital phonocardiography", "sound analysis during auscultation", "neural network", "artificial intelligence", "machine learning". R e sults . Of the 64 publications screened, 15 studies meeting the inclusion criteria were included in the final analysis. Conclusion . The effectiveness of digital technologies and artificial intelligence for the automatic diagnosis of heart murmurs has been confirmed. Further development is focused on improving the analysis of phonocardiographic signals, advancing telemedicine, and creating universal platforms for population screening.

Relevance. During immediate prosthetic rehabilitation in patients with periodontitis, the early prosthodontic stage– specifically the placement of a provisional restoration—is frequently associated with mucosal ischemic blanching, pain, and unplanned adjustments. These complications largely result from the absence of quantitative criteria defining acceptable levels of soft tissue compression. A silicone-based fit-checking technique with a predefined film thickness range allows standardized seating of provisional restorations and enhances procedural reproducibility. Objective. To improve the quality of immediate prosthetic rehabilitation in patients with periodontitis by using a silicone-based fit-checking technique with quantitative control of film thickness combined with site-specific FDM-based modification of the pontic. Materials and methods . A single-center prospective controlled study was conducted. Two independent groups of 10 patients each were formed to assess patientand clinician-reported outcomes. In the study group, provisional restorations were seated with quantitative assessment of soft tissue compression using a thin-layer condensation silicone fit-checking material with a predefined film thickness range of 60–150 μm, followed by site-specific FDM-based modification of the pontic. In the control group, restorations were adjusted using standard chairside fitting procedures without quantitative reference values. An additional cohort of 48 patients was included for zonal assessment of silicone film thickness (five zones per patient; 240 measurements in total). Data normality was evaluated using the Shapiro–Wilk test. Intergroup differences in VAS scores were analyzed using the Mann–Whitney U test. Associations between silicone film thickness and VAS scores were examined using Spearman’s rank correlation coefficient. Results. The study group demonstrated more favorable patientand clinician-reported outcomes. Median clinician VAS scores were 6.8 [6.6; 7.1] compared with 9.2 [8.9; 9.3] in the control group, while median patient VAS scores were 7.3 [5.3; 7.9] versus 8.8 [8.6; 9.0], respectively (n = 10 per group). Across all assessed zones (n = 240), the median silicone film thickness was 80 [60; 100] μm, with a mean value of 84.6 ± 50.3 μm and a range of 20–300 μm. The proportion of measurements within the predefined target range of 60–150 μm was 67.5%. Zonal analysis demonstrated the highest degree of controllability in the vestibular region (95.8% of measurements within the target range) and the lowest in the distal region (41.7%), indicating the need for targeted correction primarily in this area. Conclusion. Quantitative control of provisional restoration seating using a thin-layer silicone-based fit-checking technique combined with targeted FDM-based modification is clinically feasible, reduces perceived soft tissue trauma, and improves the reproducibility of the early prosthodontic phase of immediate rehabilitation in patients with periodontitis. The proposed numerical thresholds and zonal reference values may serve as formalized quality endpoints and provide a foundation for protocol standardization in clinical practice. Multicenter randomized controlled trials are required to evaluate mediumand long-term outcomes.

Background — Chronic heart failure (CHF) is one of the leading causes of mortality. Many factors can influence the risk of mortality in patients with CHF. Therefore, it is necessary to clarify the predictors of mortality in patients with CHF. The goal of our study was to identify predictors of adverse prognosis in patients with CHF. Methods — The study included 591 patients with CHF at 60 medical facilities registered in Samara Region CHF Registry during one month in 2022. Their median age was 71.0 (64.0-80.0) years, and 339 (57.4%) of them were men. The follow-up period lasted 18 months, during which 198 (33.5%) patients died. Results — According to the results of multivariate analysis, prognostic factors associated with mortality in patients with CHF were age (OR 1.024, 95% confidence interval [CI] 1.007-1.042, p=0.006), NYHA functional class IV (OR 2.226, 95% CI 1.358-3.649, p=0.002), pleural effusion (OR 1.423, 95% CI 0.973-2.083, p=0.069), oxygen therapy in outpatient settings (2.401, 95% CI 0.963-5.988, p=0.06), inotropic therapy in a hospital settings (OR 1.559, 95% CI 0.924-2.630, p=0.096), and left ventricular ejection fraction (LVEF) < 40% (OR 1.580, 95% CI 1.066-2.342, p=0.023). Previous cardiac surgery was inversely associated with the probability of death (OR 0.476, 95% CI 0.3-0.755, p=0.002). The area under the ROC curve, corresponding to the relationship between mortality and the value of the logistic regression function, was 0.691 (p=0.023) (95% CI 0.645-0.736). Conclusion — Predictors of mortality in patients with CHF over an 18-month follow-up period include age, NYHA functional class IV CHF, pleural effusion, oxygen therapy in outpatient settings, inotropic therapy in a hospital setting, and LVEF < 40%. Previous cardiovascular surgeries had a favorable effect on mortality in these patients.

Aim   Analysis of literature on the effect of telemedicine technologies on the adherence of patients with chronic heart failure (CHF) to therapy at the outpatient stage of treatment. Material and methods Publications were searched for in PubMed, Web of Science, Scopus, CyberLeninka, and Google Scholar using the key terms "telemedicine technologies," "remote monitoring," "commitment to treatment," "mobile applications in treatment," "telemedicine," and "telemonitoring." Reports were selected (January 2015-January 2025) independently by two researchers who reached a consensus. Based on the selection of literature sources, four studies that met the search criteria were included in the meta-analysis. The analysis was performed with the Review Manager 5.4.1 and Comprehensive Meta-Analysis 3.0 software using the fixed-effects model (inverse variance method). The results were visualized using a forest plot. Heterogeneity assessment included the Q-test and the I2 index (criterion: Results   The overall adherence rate in the reviewed studies was 64.8% (79/122 patients) in the telemonitoring group and 59.2% (74/125 patients) in the standard observation group, for a total of 61.9% (153/247 patients). Thus, the relative risk of treatment adherence was 1.10 (95% confidence interval (CI) 0.91-1.33; p=0.33). When assessing the homogeneity of the studies with the Pearson’s chi-square test, a statistically non-significant result was obtained (Pearson’s chi-square 2.74 with a degree of freedom of 2, p=0.25), and the heterogeneity index I2=27%, which determines the low heterogeneity of the studies. To identify systematic errors in the studies that used telemonitoring and the standard approach, funnel plot data were accounted for, which indicated an asymmetric distribution of data. Conclusion Telemonitoring with mobile apps and wearable devices is beneficial for patient adherence to treatment. Active implementation of these technologies is required to improve the monitoring of the use of prescribed therapy by patients with CHF.