Medical Decision Support System Research Articles

The Effect of an Electronic Medical Record-Based Clinical Decision Support System on Adherence to Clinical Protocols in Inflammatory Bowel Disease Care: Interrupted Time Series Study.

Clinical decision support systems (CDSSs) embedded in electronic medical records (EMRs), also called electronic health records, have the potential to improve the adoption of clinical guidelines. The University of Alberta Inflammatory Bowel Disease (IBD) Group developed a CDSS for patients with IBD who might be experiencing disease flare and deployed it within a clinical information system in 2 continuous time periods. This study aims to evaluate the impact of the IBD CDSS on the adherence of health care providers (ie, physicians and nurses) to institutionally agreed clinical management protocols. A 2-period interrupted time series (ITS) design, comparing adherence to a clinical flare management protocol during outpatient visits before and after the CDSS implementation, was used. Each interruption was initiated with user training and a memo with instructions for use. A group of 7 physicians, 1 nurse practitioner, and 4 nurses were invited to use the CDSS. In total, 31,726 flare encounters were extracted from the clinical information system database, and 9217 of them were manually screened for inclusion. Each data point in the ITS analysis corresponded to 1 month of individual patient encounters, with a total of 18 months of data (9 before and 9 after interruption) for each period. The study was designed in accordance with the Statement on Reporting of Evaluation Studies in Health Informatics (STARE-HI) guidelines for health informatics evaluations. Following manual screening, 623 flare encounters were confirmed and designated for ITS analysis. The CDSS was activated in 198 of 623 encounters, most commonly in cases where the primary visit reason was a suspected IBD flare. In Implementation Period 1, before-and-after analysis demonstrates an increase in documentation of clinical scores from 3.5% to 24.1% (P<.001), with a statistically significant level change in ITS analysis (P=.03). In Implementation Period 2, the before-and-after analysis showed further increases in the ordering of acute disease flare lab tests (47.6% to 65.8%; P<.001), including the biomarker fecal calprotectin (27.9% to 37.3%; P=.03) and stool culture testing (54.6% to 66.9%; P=.005); the latter is a test used to distinguish a flare from an infectious disease. There were no significant slope or level changes in ITS analyses in Implementation Period 2. The overall provider adoption rate was moderate at approximately 25%, with greater adoption by nurse providers (used in 30.5% of flare encounters) compared to physicians (used in 6.7% of flare encounters). This is one of the first studies to investigate the implementation of a CDSS for IBD, designed with a leading EMR software (Epic Systems), providing initial evidence of an improvement over routine care. Several areas for future research were identified, notably the effect of CDSSs on outcomes and how to design a CDSS with greater utility for physicians. CDSSs for IBD should also be evaluated on a larger scale; this can be facilitated by regional and national centralized EMR systems.

Extensive Review of Literature on Explainable AI (XAI) in Healthcare Applications

Abstract: Artificial Intelligence (AI) techniques are widely being used in the medical fields or various applications including diagnosis of diseases, prediction and classification of diseases, drug discovery, etc. However, these AI techniques are lacking in the transparency of the predictions or decisions made due to their black box-type operations. The explainable AI (XAI) addresses such issues faced by AI to make better interpretations or decisions by physicians. This article explores XAI techniques in the field of healthcare applications, including the Internet of Medical Things (IoMT). XAI aims to provide transparency, accountability, and traceability in AI-based systems in healthcare applications. It can help in interpreting the predictions or decisions made in medical diagnosis systems, medical decision support systems, smart wearable healthcare devices, etc. Nowadays, XAI methods have been utilized in numerous medical applications over the Internet of Things (IOT), such as medical diagnosis, prognosis, and explanations of the AI models, and hence, XAI in the context of IoMT and healthcare has the potential to enhance the reliability and trustworthiness of AI systems.

ORGANIZATION OF DISPENSARY OBSERVATION OF PATIENTS WITH PATHOLOGY OF THE MACULAR RETINA USING ARTIFICIAL INTELLIGENCE SYSTEMS

Background: Despite the mention in the provision of medical care for diseases of the organ of vision, adnexa and orbit about equipping the medical consultative and diagnostic department of the clinic with an optical coherence tomograph, dynamic monitoring of patients with retinal pathology after the start of treatment is most often carried out in a medical ophthalmological center, which reduces accessibility for patients with primary (first identified) morbidity requiring initiation of treatment. The required technology needs to be changed, intensified, incl. using artificial intelligence technologies.&#x0D; Aim: development of methodological foundations for organizational technology for dispensary observation of patients with pathology of the posterior segment of the eye using clinical decision support system (CDSS) based on artificial intelligence.&#x0D; Materials and methods: The assessment of the existing regulatory framework was carried out on the basis of an analysis of the Constitution of the Russian Federation, Federal laws, by-laws, and judicial practice. The creation of a methodology for a structured medical document describing an optical coherence tomography image was carried out using an expert method: a survey of 100 ophthalmologists with the appropriate level of education, incl. additional professional, engaged in the provision of medical services - specialized medical care (treatment) to patients with pathology of the posterior segment of the eye.&#x0D; Results: Using an expert method, 123 binary signs were selected and described to describe the structure of the macular region of the retina in normal and pathological conditions, of which 26 signs were identified that can be interpreted as predictors of worsening of the clinical course of the disease.&#x0D; Conclusion: The developed classifier made it possible to create and train a medical decision support system based on 60,000 medical images, which, as an information service, without making a diagnosis, can change the organization of the dynamic observation process: the formation of patient routing - a primary service using the developed CDSS; if there are signs of deterioration in the clinical picture - routing to a medical ophthalmological center to assess the dynamics, provide specialized services, incl. high-tech medical care.

Bridging the voice of healthcare to digital transformation in practice – a holistic approach

BackgroundDigital transformation is key for healthcare to meet future needs and expectations and compete equally with new actors on the market. Medical digital diagnosis tools and clinical decision support systems (CDSS) are improving and offer new opportunities. To introduce new technology in healthcare can however be a challenging mission, too often ending in failure, with excessive costs or the actual transformation work not being carried out at all. It is unclear how to drive the establishment to reach desired results in this environment, and how industrial experiences can be used to support healthcare.ObjectiveThe objective of this study was to develop a holistic approach for introducing new information technology (IT), such as a CDSS, into a primary healthcare organization supported by industry best practices for digital transformation.MethodsThis qualitative study used a combined inductive and deductive method where the perceptions and beliefs of selected primary healthcare stakeholders were used as directions for developing an approach that could utilize existing industry best practices for digital transformation.ResultsA holistic healthcare-ified approach including 20 requirements was developed, that meet the needs of healthcare. The voice of healthcare was used as prism to healthcare-ify the industry practices and adapt it to match specific healthcare conditions. An example was provided showing how the research could be put into practice.ConclusionsThis study proposed a holistic approach, based on industry best practices, but adapted to healthcare using the voice of healthcare as a bridge, that may be used to introduce CDSS and other IT into a primary healthcare organization and step up the needed digitalization.

Ethical considerations in AI-enhanced medical decision support systems: A review

As Artificial Intelligence (AI) continues to play an increasingly pivotal role in medical decision support systems, the ethical implications of its integration into healthcare practices demand comprehensive examination. This review delves into the ethical considerations surrounding AI-enhanced medical decision support systems, aiming to provide insights into the challenges, existing frameworks, exemplary practices, and emerging trends in this rapidly evolving field. The significance of ethical considerations is underscored by the patient-centric focus, emphasizing the impact of AI on patient outcomes and the delicate balance between technological advancements and patient welfare. Trust and transparency emerge as critical pillars, exploring the role of trust in medical decision-making and the imperative of ensuring transparency in AI algorithms to foster confidence among healthcare professionals and patients. Ethical challenges, including privacy and confidentiality concerns, biases in AI algorithms, and issues related to informed consent, are thoroughly examined. Strategies for safeguarding patient data, mitigating biases, and transparently communicating with patients are explored to address these challenges. The role of accountability and responsibility is delineated, defining the ethical responsibilities of both healthcare professionals and AI developers. The review surveys existing ethical frameworks in healthcare AI and evaluates their applicability and effectiveness. Additionally, it highlights recent proposals for ethical guidelines, emphasizing the need to integrate ethical considerations into the entire development life cycle of AI-enhanced medical decision support systems. Case studies and exemplary practices from healthcare institutions implementing ethical AI serve to illustrate real-world applications and offer insights into best practices. The evolving landscape of ethical AI research is explored, showcasing ongoing initiatives and potential innovations that hold promise for addressing ethical challenges in the future. This review underscores the paramount importance of ethical considerations in the integration of AI into medical decision support systems. It provides a comprehensive overview of current challenges, existing frameworks, exemplary practices, and emerging trends, emphasizing the ongoing need for vigilance and ethical governance to ensure the responsible and beneficial deployment of AI in healthcare.

Analyzing the impact of data visualization applications for diagnosing the health conditions through hesitant fuzzy-based hybrid medical expert system

Effectively managing healthcare data is crucial for accurate diagnosis and personalized patient care. As the utilization of healthcare data grows for personalized care, concerns about reliability, privacy, and security have emerged. To address these issues, this research explores the fusion of analytical techniques with interactive visual representations, known as visual analytics, as a promising solution. The focus is on evaluating the trustworthiness of healthcare data in Kingdom of Saudi Arabia, particularly the capability of visual analytics tools in facilitating accurate and secure healthcare data analysis. This study tackles challenges such as the absence of specific evaluation criteria, the need to process vast healthcare datasets, the establishment of trust, and the necessity for automation. In response, a hybrid medical decision support system is introduced, leveraging hesitant fuzzy decision systems. The primary objective is to evaluate trustworthiness of visual analytics tools for disease diagnosis within healthcare data. Within the framework of hesitant fuzzy logic, the paper employs a medical multi-criteria decision-making system that integrates the analytic network process and the technique for order of preference by similarity to an ideal solution. Rigorous validation ensures the accuracy and reliability of the findings. The research not only provides valuable insights but also conducts comparative analyses of the proposed models against existing ones, demonstrating the practicality of optimal decision-making in Saudi Arabia environment of healthcare scenarios. Several popular alternatives of healthcare based tools have been used in this study such as Tableau, JupyteR, Zoho Reports, QlikView, Visual.ly, DOMO BI, SAS Visual Analytics. From the results achieved DOMO BI visual analytics tool is found to be most secure and robust tool for healthcare professionals. This effort aims to enhance patient care and outcomes, ultimately contributing to the improvement of the overall healthcare landscape in Saudi Arabia.

Enhancing Medical Decision Making: A Semantic Technology-Based Framework for Efficient Diagnosis Inference

In the dynamic landscape of healthcare, decision support systems (DSS) confront continuous challenges, especially in the era of big data. Background: This study extends a Q&amp;A-based medical DSS framework that utilizes semantic technologies for disease inference based on a patient’s symptoms. The framework inputs “evidential symptoms” (symptoms experienced by the patient) and outputs a ranked list of hypotheses, comprising an ordered pair of a disease and a characteristic symptom. Our focus is on advancing the framework by introducing ontology integration to semantically enrich its knowledgebase and refine its outcomes, offering three key advantages: Propagation, Hierarchy, and Range Expansion of symptoms. Additionally, we assessed the performance of the fully implemented framework in Python. During the evaluation, we inspected the framework’s ability to infer the patient’s disease from a subset of reported symptoms and evaluated its effectiveness in ranking it prominently among hypothesized diseases. Methods: We conducted the expansion using dedicated algorithms. For the evaluation process, we defined various metrics and applied them across our knowledge base, encompassing 410 patient records and 41 different diseases. Results: We presented the outcomes of the expansion on a toy problem, highlighting the three expansion advantages. Furthermore, the evaluation process yielded promising results: With a third of patient symptoms as evidence, the framework successfully identified the disease in 94% of cases, achieving a top-ranking accuracy of 73%. Conclusions: These results underscore the robust capabilities of the framework, and the enrichment enhances the efficiency of medical experts, enabling them to provide more precise and informed diagnostics.

Application of virtual reality technology in medical decision support systems

Application of virtual reality technology in medical decision support systems

Technologies of Bioimpedance Spectroscopy in Decision Support Systems for the Diagnosis of Socially Significant Diseases

The purpose of the research is to develop methods for the synthesis of hybrid classifiers to assess the risk of socially significant diseases using bioimpedance analysis.Methods. We developed a descriptor approach using impedance spectroscopy results, generating four amplitudephase-frequency responses from four quasi-orthogonal leads. They create the feature spaces necessary for our hybrid classifier in the diagnosis of pancreatic diseases, the autonomous intelligent agents of which are built on various paradigms: probabilistic neural networks, fuzzy logical inference, fully connected feedforward neural networks. We also presented a device structure for creating an informative feature space.Results. Experimental studies of the proposed methods and means of classifying medical rice were carried out on diagnostic tasks according to the classes "acute destructive pancreatitis" – "no acute destructive pancreatitis" and differential diagnosis tasks according to the classes "prostate cancer" ‒ "chronic pancreatitis". They showed that incorporating multi-frequency sensing into neural network-based classifiers allows the development of clinical decision support systems for disease diagnosis that are comparable in performance to existing clinical diagnostic methods. The results were confirmed in groups of male and female patients at different stages of cancer aged 25 to 80 years using a variety of diagnostic methods, including history, physical examination, assessment of comorbidities, laboratory tests, ultrasound, laparoscopy, intraoperative exploration and computed tomography.Conclusion. The use of bioimpedance spectroscopy and hybrid classifier models opens up new opportunities for accessible and objective diagnosis of pancreatic diseases, expanding the capabilities of intelligent medical decision support systems.

Problematic issues and development of classifications of the main parameters of quality and adherence to pharmacotherapy. Part II: physician adherence to clinical guidelines and official drug labels

The second publication is presented, devoted to the classification of the main parameters of quality and adherence to pharmacotherapy. Despite the fact that patients play a central role in the problem of medical adherence, the role of attending physicians is extremely significant both in prescribing/non-prescribing proven effective and safe therapy, and in influencing patient medical adherence. This led to the identification of a special commitment type — the physician adherence to the basic principles of rational treatment. Collection, analysis and systematization of information, as well as the related development of a classification of physician adherence/non-adherence were the aim of this review and analytical work. We searched data using keywords among Russian and English-­language sources. When compiling the classification, categories of adherence/non-adherence were identified in accordance with the implementation by physicians of the main provisions of clinical guidelines for long-term therapy for patients with non-communicable diseases (NCDs), as well as following official drug labels reflecting data on clinical pharmacology, the main side effects, drug-drug interactions. The classification of physician adherence/non-adherence to implementing the principles of rational treatment of patients with NCDs can be used in the development of electronic medical decision support systems for providing therapy with proven effectiveness and safety in specific clinical situations. In addition, on the basis of the developed classification, checklists for monitoring by healthcare specialists of the implementation of the rational treatment of patients with chronic NCDs should be created.

A Cognitive Medical Decision Support System for IoT-Based Human-Computer Interface in Pervasive Computing Environment

A Cognitive Medical Decision Support System for IoT-Based Human-Computer Interface in Pervasive Computing Environment

Risk factor determinants and comparison of supervised machine learning algorithms to predict head and neck cancer

Head and Neck Cancer (HNC) has emerged as a major public health concern in India. The Indian state Bihar, which ranks fourth in the total number of new cancer cases, head and neck cancer is the second most common cancer and has an increasing trend. The application of machine learning (ML) in disease diagnosis is increasing gradually. With this background, an attempt has been made to determine the risk factors and compare the performance of different variants of supervised ML algorithms for HNC prediction. The study confirms that poor oral hygiene, tobacco, alcohol and human papilloma virus (HPV) infections are the significant risk factor for HNC occurrence in Bihar. In comparison to all the variants of supervised machine learning algorithm, Random Forest showed maximum accuracy. This study will beneficial in medical decision support systems.

Модель оценки значимости побочных эффектов при применении нескольких лекарственных средств

The model for ranking side effects in the combined use of drugs is constructed using the example of chronic heart failure. A numerical algorithm based on the allocation of fully connected subgraphs has been developed, which reduces the amount of calculations when analyzing combinations of several drugs. The results of the test calculations are presented. The program being developed can be useful as a medical decision support system.

ОБЗОР МЕТОДИЧЕСКИХ ПОДХОДОВ К ОЦЕНКЕ КАЧЕСТВА ВЕДЕНИЯ ЭЛЕКТРОННЫХ МЕДИЦИНСКИХ КАРТ

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 realworld 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.

ТЕХНОЛОГИИ ИСКУССТВЕННОГО ИНТЕЛЛЕКТА В ОЦЕНКЕ ПЕРФУЗИИ МИОКАРДА С ПОМОЩЬЮ ПОЗИТРОННОЙ ЭМИССИОННОЙ ТОМОГРАФИИ С ИСПОЛЬЗОВАНИЕМ 82Rb-ХЛОРИДА

At the moment, one of the most common causes of morbidity and mortality is coronary heart disease, which determines the need to develop methods for its diagnosis. Among diagnostic methods, non-invasive methods occupy a special place, in particular, determination of myocardial perfusion. One of the “gold standards” for assessing cardiac muscle perfusion is positron emission tomography combined with computed tomography (PET/CT) with 82Rb-chloride. Recently, attempts have been actively made to introduce the use of artificial intelligence in a variety of areas of medical clinical practice, including the development of medical decision support systems, as well as neural networks for assessing the results of diagnostic studies. In particular, there is information about attempts to use artificial intelligence in assessing myocardial perfusion using PET/CT with 82Rb-chloride. This paper analyzes the possibilities and prospects for using artificial intelligence in assessing the results of PET/CT with 82Rb-chloride. The use of well-trained neural networks and machine learning algorithms can significantly increase the accuracy of diagnosing coronary heart disease by improving the quality of images, analyzing the data obtained, or calculating characteristics and indicators, the quantitative interpretation of which may be difficult for a doctor. Neural networks are able to take into account in the prognosis both clinical and anamnestic data and additional parameters determined from research data, which the doctor may not pay attention to, which determines the relevance and prospects for the use of artificial intelligence in relation to the interpretation of 82Rb-PET/CT results.

Development of a testing machine for biaxial testing of soft tissue and biomaterials

Knowledge of the mechanical properties of soft tissues and biomaterials is the keystone for predictive modeling in cardio- and endovascular surgery, traumatology and orthopedics, neurosurgery, and the development of tissue-engineering constructs. Needed mechanical properties can be only explored via experimental tests. Therefore, mechanical experiments are forming the foundation of medical decision support systems, which are impossible without adequate mathematical models of the processes involved. Sirius University of Science and Technology has launched the project "Experimental research and mathematical modeling of the mechanical behavior of biomaterials". The goal of the project is to create an infrastructure for experimental study of the mechanics of soft tissues and biomaterials. At this stage, a machine for biaxial testing of soft tissues and biomaterials has been developed and put into trial operation. The developed machine takes into account the features of working with soft tissues: small sample size, maintaining certain environmental conditions to preserve the sample, various types of sample gripping to minimize edge effects and no slippage, non-contact methods for assessing displacement and deformation fields, variability in sample loading protocols. The result of the test is a data set that includes a test report and synchronized output readings of strain gauge, temperature, optoelectronic sensors with high-resolution images. Uniaxial verification tests of acrylic tapes VHB4910 were carried out by the developed machine. The results were compared with data obtained on a commercial dynamic mechanical analyzer RSA-G2 (TA Instruments) for similar materials under the same test conditions. A series of biaxial cyclic experiments were carried out for the pericardium of cattle, the cardiovascular patch Ecofolon PS04/PS01 (JSC NPK Ecoflon, St. Petersburg, Russia), and the GoreTex vascular graft (W.L. Gore and Associates, USA).

Survival probability function as a supporting element in automated triage system on the battlefield

Mass events are incidents involving a large number of casualties. They are an integral part of combat operations. These events are characterized by exceeding the capabilities of the rescuers present at the scene during a specific phase of the operation. The difference in triage applied in combat compared to triage used in civilian mass events is directly related to their specific nature and intended objectives. The article presents an innovative segregation algorithm that takes into account the value of the so-called "survival function," which is a component of a medical evacuation decision support system based on the integration of monitoring and analysis of soldier's vital parameters with the medical security system.

Designing Digital Therapeutic Content Using Chronic Disease Data: A Focus on Improving Urinary Dysfunction.

In recent years, advancements in information and communication technologies, including artificial intelligence, big data, virtual reality, and augmented reality, have driven substantial growth in the field of digital medical diagnosis and treatment, thereby enhancing quality of life. Beginning in the mid-2010s with the advent of digital healthcare applications, and further accelerated by the impact of coronavirus disease 2019, digital therapeutic products have profoundly influenced society. Nevertheless, the expansion of digital therapeutics has encountered challenges associated with regulatory hurdles, differentiation from general digital healthcare, and the necessity for trustworthiness, which have contributed to a slower rate of progress. This study proposes a 3P content model-encompassing pre-education, prediction/diagnosis/treatment, and postmanagement-to increase the trustworthiness of digital therapeutics. The design of the 3P content model includes a fundamental structure that establishes networks with healthcare institutions, aiming to increase the reliability of data utilization and to facilitate integration with medical decision support systems. For case development, the study introduces a prototype of a mobile application that utilizes chronic disease urinary dysfunction data, demonstrating the cyclical structure inherent in the 3P content model.

Development and Evaluation of Intelligent Medical Decision Support Systems

This paper investigates the development and evaluation of intelligent medical decision support systems. Firstly, the current status of such systems is introduced, including technological advancements, clinical applications, and challenges faced. Subsequently, the overall framework design, key module design and implementation, system integration, and debugging are elaborated in detail. Finally, using the publicly available MIMIC medical dataset, a comprehensive evaluation and analysis of the system are conducted in a high-performance environment. The results indicate that the optimized system can achieve high accuracy and AUC. However, there is a slight decrease in effectiveness when dealing with rare diseases, and occasional false positives in certain high-risk situations. Overall, the research confirms the promising application prospects of such systems and provides a framework and path for future development.

Rare Category Analysis for Complex Data: A Review

Though the sheer volume of data that is collected is immense, it is the rare categories that are often the most important in many high-impact domains, ranging from financial fraud detection in online transaction networks to emerging trend detection in social networks, from spam image detection on social media platforms to rare disease diagnosis in medical decision support systems. The unique challenges of rare category analysis include (1) the highly skewed class distribution; (2) the non-separable nature of the rare categories from the majority classes; (3) data and task heterogeneity; and (4) the time-evolving property of the input data sources. This survey reviews state-of-the-art techniques used in complex rare category analysis, where the majority classes have a smooth distribution while the minority classes exhibit the compactness property in the feature space or subspace. Rare category analysis aims to identify, characterize, represent, and interpret anomalies that not only show statistical significance but also exhibit interesting patterns (e.g., compactness, high-order structures, showing in a burst). We introduce our study, define the problem setting, and describe the unique challenges of complex rare category analysis. We then present a comprehensive review of recent advances that are designed for this problem setting, from rare category exploration without any label information to rare category exposition that characterizes rare examples with a compact representation, from the representation of rare patterns in a salient embedding space to the interpretation the prediction results and providing relevant clues for the end-users’ interpretation. Finally we discuss potential challenges and shed light on the future directions for complex rare category analysis. 1