The ongoing transformation of health systems around the world aims at personalized, preventive, predictive, participative precision medicine, supported by technology. It considers individual health status, conditions, and genetic and genomic dispositions in personal, social, occupational, environmental and behavioral contexts. In this way, it transforms health and social care from art to science by fully understanding the pathology of diseases and turning health and social care from reactive to proactive. The challenge is the understanding and the formal as well as consistent representation of the world of sciences and practices, i.e., of multidisciplinary and dynamic systems in variable context. This enables mapping between the different disciplines, methodologies, perspectives, intentions, languages, etc., as philosophy or cognitive sciences do. The approach requires the deployment of advanced technologies including autonomous systems and artificial intelligence. This poses important ethical and governance challenges. This paper describes the aforementioned transformation of health and social care ecosystems as well as the related challenges and solutions, resulting in a sophisticated, formal reference architecture. This reference architecture provides a system-theoretical, architecture-centric, ontology-based, policy-driven model and framework for designing and managing intelligent and ethical ecosystems in general and health ecosystems in particular. Full article

Machine learning has been proven to provide good performances on stress detection tasks using multi-modal sensor data from a smartwatch. Generally, machine learning techniques need a sufficient amount of data to train a robust model. Thus, we need to collect data from several users and send them to a central server to feed the algorithm. However, the uploaded data may contain sensitive information that can jeopardize the user’s privacy. Federated learning can tackle this challenge by enabling the model to be trained using data from all users without the user’s data leaving the user’s device. In this study, we implement federated learning-based stress detection and provide a comparative analysis between individual, centralized, and federated learning. The experiment was conducted on WESAD dataset by using Logistic Regression as the classifier. The experiment results show that in terms of accuracy, federated learning cannot reach the performance level of both individual and centralized learning. The individual learning strategy performs best with an average accuracy of 0.9998 and an average F₁-measure of 0.9996. Full article

The Italian “Istituto Superiore di Sanità” (ISS) identifies hospital-acquired infections (HAIs) as the most frequent and serious complications in healthcare. HAIs constitute a real health emergency and, therefore, require decisive action from both local and national health organizations. Information about the causative microorganisms of HAIs is obtained from the results of microbiological cultures of specimens collected from infected body sites, but microorganisms’ names are sometimes reported only in the notes field of the culture reports. The objective of our work was to build a NLP-based pipeline for the automatic information extraction from the notes of microbiological culture reports. We analyzed a sample composed of 499 texts of notes extracted from 1 month of anonymized laboratory referral. First, our system filtered texts in order to remove nonmeaningful sentences. Thereafter, it correctly extracted all the microorganisms’ names according to the expert’s labels and linked them to a set of very important metadata such as the translations into national/international vocabularies and standard definitions. As the major result of our pipeline, the system extracts a complete picture of the microorganism. Full article

In this paper, we propose a health data sharing infrastructure which aims to empower a democratic health data sharing ecosystem. Our project, named Health Democratization (HD), aims at enabling seamless mobility of health data across trust boundaries through addressing structural and functional challenges of its underlying infrastructure with the core concept of data democratization. A programmatic design of an HD platform was elaborated, followed by an introduction of one of our critical designs—a “reverse onus” mechanism that aims to incentivize creditable data accessing behaviors. This scheme shows a promising prospect of enabling a democratic health data-sharing platform. Full article

The primary goal of this paper is to develop an approach for predicting important clinical indicators, which can be used to improve treatment. Using mathematical predictive modeling algorithms, we examined the course of COVID-19-based pneumonia (CP) with inpatient treatment. Algorithms used include dynamic and ordinary Bayesian networks (OBN and DBN), popular ML algorithms, the state-of-the-art auto ML approach and our new hybrid method based on DBN and auto ML approaches. Predictive targets include treatment outcomes, length of stay, dynamics of disease severity indicators, and facts of prescribed drugs for different time intervals of observation. Models are validated using expert knowledge, current clinical recommendations, preceding research and classic predictive metrics. The characteristics of the best models are as follows: MAE of 3.6 days of predicting LOS (DBN plus FEDOT auto ML framework), 0.87 accuracy of predicting treatment outcome (OBN); 0.98 F1 score for predicting facts of prescribed drug (DBN). Moreover, the advantage of the proposed approach is Bayesian network-based interpretability, which is very important in the medical field. After the validation of other CP datasets for other hospitals, the proposed models can be used as part of the decision support systems for improving COVID-19-based pneumonia treatment. Another important finding is the significant differences between COVID-19 and non-COVID-19 pneumonia. Full article

Machine learning methods to predict the risk of epilepsy, including vascular epilepsy, in oncohematological patients are currently considered promising. These methods are used in research to predict pharmacoresistant epilepsy and surgical treatment outcomes in order to determine the epileptogenic zone and functional neural systems in patients with epilepsy, as well as to develop new approaches to classification and perform other tasks. This paper presents the results of applying machine learning to analyzing data and developing diagnostic models of epilepsy in oncohematological and cardiovascular patients. This study contributes to solving the problem of often unjustified diagnosis of primary epilepsy in patients with oncohematological or cardiovascular pathology, prescribing antiseizure drugs to patients with single seizure syndromes without finding a disease associated with these cases. We analyzed the hospital database of the V.A. Almazov Scientific Research Center of the Ministry of Health of Russia. The study included 66,723 treatment episodes of patients with vascular diseases (I10–I15, I61–I69, I20–I25) and 16,383 episodes with malignant neoplasms of lymphoid, hematopoietic, and related tissues (C81–C96 according to ICD-10) for the period from 2010 to 2020. Data analysis and model calculations indicate that the best result was shown by gradient boosting with mean accuracy cross-validation score = 0.96. f1-score = 98, weighted avg precision = 93, recall = 96, f1-score = 94. The highest correlation coefficient for G40 and different clinical conditions was achieved with fibrillation, hypertension, stenosis or occlusion of the precerebral arteries (0.16), cerebral sinus thrombosis (0.089), arterial hypertension (0.17), age (0.03), non-traumatic intracranial hemorrhage (0.07), atrial fibrillation (0.05), delta absolute neutrophil count (0.05), platelet count at discharge (0.04), transfusion volume for stem cell transplantation (0.023). From the clinical point of view, the identified differences in the importance of predictors in a broader patient model are consistent with a practical algorithm for organic brain damage. Atrial fibrillation is one of the leading factors in the development of both ischemic and hemorrhagic strokes. At the same time, brain infarction can be accompanied both by the development of epileptic seizures in the acute period and by unprovoked epileptic seizures and development of epilepsy in the early recovery and in a longer period. In addition, a microembolism of the left heart chambers can lead to multiple microfocal lesions of the brain, which is one of the pathogenetic aspects of epilepsy in elderly patients. The presence of precordial fibrillation requires anticoagulant therapy, the use of which increases the risk of both spontaneous and traumatic intracranial hemorrhage. Full article

Today, periprosthetic joint infection (PJI) is one of the leading indications for revision surgery and the most ominous complication in artificial joint patients. The current state of the art for treating PJI requires the development of methods for planning the costs at different scales to facilitate the selection of the best treatment methods. In this paper, we perform a cost-effectiveness assessment for strategies related to the treatment of PJI using a composite decision support modeling framework. Within the framework, two models are implemented: a detailed discrete-event probabilistic model based on the decision tree approach and a dynamic Markov model with generalized states. The application of the framework is demonstrated on the dataset which was provided by the Russian Scientific Research Institute of Traumatology and Orthopedics named after R.R. Vreden. The analyzed dataset contains 600 patient records divided into two groups (retrospective group, based on old records, and prospective group, based on real-time follow-up). The cost-effectiveness of treatment methods was compared based on associated costs and QALY units gained, with the mentioned two indicators calculated using two models independently from each other. As a result, two comparative rankings of cost-effectiveness of PJI treatment methods were presented based on the model output. Full article

Genomic information is a very sensitive type of digital information as it not only applies to a person, but also to close relatives. Therefore, privacy provision is key to protecting genomic information from unauthorized access. It is worth noting that most of the current genomic information formats do not provide specific mechanisms by which to secure the stored information. In order to solve, among other things, the privacy provision issue, we proposed the GIPAMS (Genomic Information Protection And Management System) modular architecture, which is based on the use of standards such as ISO/IEC 23092 and a few GA4GH (Global Alliance for Genomics and Health) initiatives. Some of the GIPAMS modules have already been implemented, mainly based on ISO/IEC 23092 features, and we are conducting work on the complete version of the architecture, and other standards are also considered. One of the objectives of GIPAMS is to enable the management of different formats of genomic information in a unique and interoperable way, providing privacy and security for formats that do not currently support them. Full article

Digital interventions for increasing physical activity behavior have shown great potential, especially those with social media. Chatbots, also known as conversational agents, have emerged in healthcare in relation to digital interventions and have proven effective in promoting physical activity among adults. The study’s objective is to explore users’ experiences with a social media chatbot. The concept and the prototype development of the social media chatbot MYA were realized in three steps: requirement analysis, concept development, and implementation. MYA’s design includes behavior change techniques effective in increasing physical activity through digital interventions. Participants in a usability study answered a survey with the Chatbot Usability Questionnaire (CUQ), which is comparable to the Systems Usability Scale. The mean CUQ score was below 68, the benchmark for average usability. The highest mean CUQ score was 64.5 for participants who thought MYA could help increase their physical activity behavior. The lowest mean CUQ score was 40.6 for participants aged between 50 and 69 years. Generally, MYA was considered to be welcoming, very easy to use, realistic, engaging, and informative. However, some technical issues were identified. A good and diversified user experience promotes prolonged chatbot use. Addressing identified issues will enhance users’ interaction with MYA. Full article

Aortic aneurysm (AA) rapture is one of the leading causes of death worldwide. Unfortunately, the diagnosis of AA is often verified after the onset of complications, in most cases after aortic rupture. The aim of this study was to evaluate the frequency of ascending aortic aneurysm (AscAA) and aortic dilatation (AD) in patients with cardiovascular diseases undergoing echocardiography, and to identify the main risk factors depending on the morphology of the aortic valve. We processed 84,851 echocardiographic (ECHO) records of 13,050 patients with aortic dilatation (AD) in the Almazov National Medical Research Centre from 2010 to 2018, using machine learning methodologies. Despite a high prevalence of AD, the main reason for the performed ECHO was coronary artery disease (CAD) and hypertension (HP) in 33.5% and 14.2% of the patient groups, respectively. The prevalence of ascending AD (>40 mm) was 15.4% (13,050 patients; 78.3% (10,212 patients) in men and 21.7% (2838 patients) in women). Only 1.6% (n = 212) of the 13,050 patients with AD knew about AD before undergoing ECHO in our center. Among all the patients who underwent ECHO, we identified 1544 (1.8%) with bicuspid aortic valve (BAV) and 635 with BAV had AD (only 4.8% of all AD patients). According to the results of the random forest feature importance analysis, we identified the eight main factors of AD: age, male sex, vmax aortic valve (AV), aortic stenosis (AS), blood pressure, aortic regurgitation (AR), diabetes mellitus, and heart failure (HF). The known factors of AD-like HP, CAD, hyperlipidemia, BAV, and obesity, were also AD risk factors, but were not as important. Our study showed a high frequency of AscAA and dilation. Standard risk factors of AscAA such as HP, hyperlipidemia, or obesity are significantly more common in patients with AD, but the main factors in the formation of AD are age, male sex, vmax AV, blood pressure, AS, AR, HF, and diabetes mellitus. In males with BAV, AD incidence did not differ significantly, but the presence of congenital heart disease was one of the 12 main risk factors for the formation of AD and association with more significant aortic dilatation in AscAA groups. Full article

To improve patient outcomes after trauma, the need to decrypt the post-traumatic immune response has been identified. One prerequisite to drive advancement in understanding that domain is the implementation of surgical biobanks. This paper focuses on the outcomes of patients with one of two diagnoses: post-traumatic arthritis and osteomyelitis. In creating surgical biobanks, currently, many obstacles must be overcome. Roadblocks exist around scoping of data that is to be collected, and the semantic integration of these data. In this paper, the generic component model and the Semantic Web technology stack are used to solve issues related to data integration. The results are twofold: (a) a scoping analysis of data and the ontologies required to harmonize and integrate it, and (b) resolution of common data integration issues in integrating data relevant to trauma surgery. Full article

In the context of the ongoing digitization of interdisciplinary subjects, the need for digital literacy is increasing in all areas of everyday life. Furthermore, communication between science and society is facing new challenges, not least since the COVID-19 pandemic. In order to deal with these challenges and to provide target-oriented online teaching, new educational concepts for the transfer of knowledge to society are necessary. In the transfer project “Zukunftslabor Gesundheit” (ZLG), a didactic concept for the creation of E-Learning classes was developed. A key factor for the didactic concept is addressing heterogeneous target groups to reach the broadest possible spectrum of participants. The concept has already been used for the creation of the first ZLG E-Learning courses. This article outlines the central elements of the developed didactic concept and addresses the creation of the ZLG courses. The courses created so far appeal to different target groups and convey diverse types of knowledge at different levels of difficulty. Full article

The use of eHealth and healthcare services are becoming increasingly common across networks and ecosystems. Identifying the quality and health impact of these services is a big problem that in many cases it is difficult determine. Health ecosystems are seldom designed with privacy and trust in mind, and the service user has almost no way of knowing how much trust to place in the service provider and other stakeholders using his or her personal health information (PHI). In addition, the service user cannot rely on privacy laws, and the ecosystem is not a trustworthy system. This demonstrates that, in real life, the user does not have significant privacy. Therefore, before starting to use eHealth services and subsequently disclosing personal health information (PHI), the user would benefit from tools to measure the level of privacy and trust the ecosystem can offer. For this purpose, the authors developed a solution that enables the service user to calculate a Merit of Service (Fuzzy attractiveness rating (FAR)) for the service provider and for the network where PHI is processed. A conceptual model for an eHealth ecosystem was developed. With the help of heuristic methods and system and literature analysis, a novel proposal to identify trust and privacy attributes focused on eHealth was developed. The FAR value is a combination of the service network’s privacy and trust features, and the expected health impact of the service. The computational Fuzzy linguistic method was used to calculate the FAR. For user friendliness, the Fuzzy value of Merit was transformed into a linguistic Fuzzy label. Finally, an illustrative example of FAR calculation is presented. Full article

The complications of thoracic aortic disease include aortic dissection and aneurysm. The risks are frequently compounded by many cardiovascular comorbidities, which makes the process of clinical decision making complicated. The purpose of this study is to develop risk predictive models for patients after thoracic aneurysm surgeries, using integrated data from different medical institutions. Seven risk features were formulated for prediction. The CatBoost classifier performed best and provided an ROC AUC of 0.94–0.98 and an F-score of 0.95–0.98. The obtained results are widely in line with the current literature. The obtained findings provide additional support for clinical decision making, guiding a patient care team prior to surgical treatment, and promoting a safe postoperative period. Full article

For maintaining collaboration and coordination among emergency department (ED) caregivers, it is essential to effectively share patient-centered information. Indirect activities on patients, such as searching for laboratory results and sharing information with scattered colleagues, waste resources to the detriment of patients and staff. Therefore, we conducted a pilot study to evaluate the initial efficacy of a mobile app to facilitate rapid mobile access to central laboratory results and remote interprofessional communication. A total of 10 ED residents and registered nurses were randomized regarding the use of the app versus conventional methods during semi-simulated scenarios in a pediatric ED (PED). The primary outcome was the elapsed time in minutes in each group from the availability of laboratory results to their consideration by participants. The secondary outcome was the elapsed time to find a colleague upon request. Time to consider laboratory results was significantly reduced from 23 min (IQR 10.5–49.0) to 1 min (IQR 0–5.0) with the use of the app compared to conventional methods (92.2% reduction in mean times, p = 0.0079). Time to find a colleague was reduced from 24 min to 1 min (i.e., 93.0% reduction). Dedicated mobile apps have the potential to improve information sharing and remote communication in emergency care. Full article

Electronic medical records (EMRs) include many valuable data about patients, which is, however, unstructured. Therefore, there is a lack of both labeled medical text data in Russian and tools for automatic annotation. As a result, today, it is hardly feasible for researchers to utilize text data of EMRs in training machine learning models in the biomedical domain. We present an unsupervised approach to medical data annotation. Syntactic trees are produced from initial sentences using morphological and syntactical analyses. In retrieved trees, similar subtrees are grouped using Node2Vec and Word2Vec and labeled using domain vocabularies and Wikidata categories. The usage of Wikidata categories increased the fraction of labeled sentences 5.5 times compared to labeling with domain vocabularies only. We show on a validation dataset that the proposed labeling method generates meaningful labels correctly for 92.7% of groups. Annotation with domain vocabularies and Wikidata categories covered more than 82% of sentences of the corpus, extended with timestamp and event labels 97% of sentences got covered. The obtained method can be used to label EMRs in Russian automatically. Additionally, the proposed methodology can be applied to other languages, which lack resources for automatic labeling and domain vocabulary. Full article