Medical Information Processing Research Articles

ВЕЛИКІ МОВНІ МОДЕЛІ ТА ОСОБИСТА ІНФОРМАЦІЯ: ПРОБЛЕМИ БЕЗПЕКИ ТА ШЛЯХИ ЇХ ВИРІШЕННЯ ЗА ДОПОМОГОЮ АНОНІМІЗАЦІЇ

In light of the growing capabilities of Large Language Models (LLMs), there is an urgent need for effective methods to protect personal data in online texts. Existing anonymization methods often prove ineffective against complex LLM analysis algorithms, especially when processing sensitive information such as medical data. This research proposes an innovative approach to anonymization that combines k-anonymity and adversarial methods. Our approach aims to improve the efficiency and speed of anonymization while maintaining a high level of data protection. Experimental results on a dataset of 10,000 comments showed a 40% reduction in processing time (from 250 ms to 150 ms per comment) compared to traditional adversarial methods, a 5% improvement in medical data anonymization accuracy (from 90% to 95%), and a 7% improvement in data utility preservation (from 85% to 92%). Special attention is paid to the application of the method in the context of interaction with LLM-based chatbots and medical information processing. We conduct an experimental evaluation of our method, comparing it with existing industrial anonymizers on real and synthetic datasets. The results demonstrate significant improvements in both data utility preservation and privacy protection. Our method also takes into account GDPR requirements, setting a new standard in the field of data anonymization for AI interactions. This research offers a practical solution for protecting user privacy in the era of LLMs, especially in sensitive areas such as healthcare. Keywords: AI, data security, ML, LLM, privacy.

Acoustic metagrating focusing and Bessel vortexes

Acoustic focusing and Bessel vortexes have great potential in medical ultrasound, particle trapping, and information processing. Based on the generalized Snell's law (GSL), metasurface focusing and Bessel vortexes were achieved by using in-plane phase profiles to shape wavefronts. Recent developments in acoustic metagratings (AMs) have demonstrated an extension of the GSL capable of switching transmitted and reflected vortexes that are determined by the parity of the number of wave propagation trips. However, these metagratings were designed with a certain one-dimensional phase gradient along the azimuthal direction, and the propagation of vortexes were generally fixed into cylindrical waveguides owing to energy divergence. The propagation and manipulation of acoustic vortexes in three-dimensional (3D) free space, caused by AMs with two-dimensional (2D) aperiodic phase gradients, still pose a great challenge. Here, we experimentally demonstrate two types of switchable acoustic lenses with focusing and Bessel vortexes. Based on the GSL extension, by separating and attaching 2D dual-layer aperiodic AMs in both lenses, the switch between the reflected focusing vortex and transmitted focusing/Bessel vortex with the same focus length and topological charge in 3D free space can be observed. The designed dual-layer AMs can realize the short-range and long-range focusing vortexes in 3D free space and also have the advantage of convenient function switching, which may pave the way for designing switchable focusing vortex lenses with practical applications.

DKA-RG: Disease-Knowledge-Enhanced Fine-Grained Image–Text Alignment for Automatic Radiology Report Generation

Automatic radiology report generation is a task that combines artificial intelligence and medical information processing, and it fully relies on computer vision and natural language processing techniques. Nowadays, automatic radiology report generation is still a very challenging task because it requires semantically adequate alignment of data from two modalities: radiology images and text. Existing approaches tend to focus on coarse-grained alignment at the global level and do not take into account the disease characteristics of radiology images at fine-grained semantics, which results in the generated reports potentially omitting key disease diagnostic descriptions. In this work, we propose a new approach, disease-knowledge-enhanced fine-grained image–text alignment for automatic radiology report generation (DKA-RG). The method combines global and disease-level alignment, thus facilitating the extraction of fine-grained disease features by the model. Our approach also introduces a knowledge graph to inject medical domain expertise into the model. Our proposed DKA-RG consists of two training steps: the image–report alignment stage and the image-to-report generation stage. In the alignment stage, we use global contrastive learning to align images and texts from a high level and also augment disease contrastive learning with medical knowledge to enhance the disease detection capability. In the report generation stage, the report text generated from the images is more accurate in describing the disease information thanks to sufficient alignment. Through extensive quantitative and qualitative experiments on two widely used datasets, we validate the effectiveness of our DKA-RG on the task of radiology report generation. Our DKA-RG achieves superior performance on multiple types of metrics (natural language generation and clinical efficacy metrics) compared to existing methods, demonstrating that the method can improve the reliability and accuracy of automatic radiology report generation systems.

A Formative Usability Study Evaluating a Decision Aid for Caregivers of Incapacitated Patients

Families making medical decisions for an incapacitated loved one need to process medical information for various care pathways while balancing different perspectives and experiencing distress. A decision aid tool to help family members make medical decisions on the patient’s behalf should be easy to use and not create additional burdens. A formative usability study with high, medium, and low verbosity versions of an initial decision aid was conducted with university students before clinical testing. The results showed no significant differences in usability metrics between different verbosity levels, but the qualitative findings indicated areas of improvement in the organization of information that could lead to improved usability of the decision aid. While the findings are limited due to the study being void of some of the burdens families face in true clinical situations, the study allowed us to identify numerous usability issues before testing with target users.

Limited health literacy does not adversely affect compliance with postoperative restrictions, 90-day emergency department return, or opioid use following shoulder arthroscopy

BackgroundLimited health literacy can negatively impact how patients process medical information, make medical decisions, and navigate the healthcare system. The literature with regards to health literacy and its impact on both postoperative compliance and healthcare utilization remains scant. MethodsWe retrospectively analyzed the records for patients who underwent elective shoulder arthroscopy with a minimum 90-day follow-up at a single academic institution. Demographic data including age, gender, prior ipsilateral shoulder arthroscopy, body mass index (BMI) and age-adjusted Charlson Comorbidity Index (ACCI) were collected. A validated 9-item Literacy in Musculoskeletal Problems (LiMP) questionnaire to assess musculoskeletal health literacy (MHL) was administered preoperatively. Postoperative compliance with therapy and surgeon-directed immobilization restrictions, 90-day return to ED, and the number of opioid prescriptions filled within 3 months postoperatively was recorded. ResultsThere were 252 cases included in this study. Seventy-seven (31%) patients demonstrated adequate MHL (AMHL). On multivariable analysis, limited MHL (LMHL) was not significantly associated with 90-day postoperative ED return, compliance with postoperative surgeon instructions regarding shoulder motion or therapy restrictions, or obtaining ≥ 2 postoperative opioid prescriptions. ConclusionsLMHL is highly prevalent amongst patients undergoing elective shoulder arthroscopy. The lack of association between LMHL and postoperative compliance, 90-day ED return, or filling ≥ 2 postoperative opioid prescriptions suggests that further research is needed to identify more relevant modifiable risk factors that could reduce these negative clinical outcomes and healthcare utilization patterns.

Cost analysis study comparing the impact of treatment with aprotinin versus tranexamic acid in cardiac surgery under cardiopulmonary bypass

An increased risk of mortality and postoperative side effects led to aprotinin (Trasylol®) withdrawal from the market in 2008, but since 2018 aprotinin has again been used in France. The French retrospective multicentre APACHE study (AProtinin versus tranexamic Acid in Cardiac surgery patients with High-risk for Excessive bleeding) compared the efficacy of tranexamic acid versus half-dose aprotinin. The aim of this study, ancillary to the APACHE study, is to carry out a medico-economic analysis of the use of these two antifibrinolytics on an APACHE subpopulation. Economic data on reimbursement by the French health insurance system were extracted from the program for the data processing of medical information, and quantitative data on the cost of healthcare products were obtained from the hospital pharmacy software. The main analysis of costs for the population shows that the global valuation was not significantly different between the two treatment groups (P=0.60), but the costs of blood products included in the related hospital stay group (Groupe Homogène de séjour [GHS]) (whole blood, platelets and plasma) were higher for the tranexamic acid group (P=0.007). In a sub-analysis of patients alive at discharge, the costs of blood products in addition to GHS (blood-derived medicines) and the costs of blood products in the GHS were higher for the tranexamic acid group (P=0.04 and 0.001, respectively). The additional cost of aprotinin at the time of purchase is offset by the additional costs of blood products in the tranexamic acid group.

Adaption BERT for Medical Information Processing with ChatGPT and Contrastive Learning

Adaption BERT for Medical Information Processing with ChatGPT and Contrastive Learning

Performance of ChatGPT on the Taiwan urology board examination: insights into current strengths and shortcomings.

To compare ChatGPT-4 and ChatGPT-3.5's performance on Taiwan urology board examination (TUBE), focusing on answer accuracy, explanation consistency, and uncertainty management tactics to minimize score penalties from incorrect responses across 12 urology domains. 450 multiple-choice questions from TUBE(2020-2022) were presented to two models. Three urologists assessed correctness and consistency of each response. Accuracy quantifies correct answers; consistency assesses logic and coherence in explanations out of total responses, alongside a penalty reduction experiment with prompt variations. Univariate logistic regression was applied for subgroup comparison. ChatGPT-4 showed strengths in urology, achieved an overall accuracy of 57.8%, with annual accuracies of 64.7% (2020), 58.0% (2021), and 50.7% (2022), significantly surpassing ChatGPT-3.5 (33.8%, OR = 2.68, 95% CI [2.05-3.52]). It could have passed the TUBE written exams if solely based on accuracy but failed in the final score due to penalties. ChatGPT-4 displayed a declining accuracy trend over time. Variability in accuracy across 12 urological domains was noted, with more frequently updated knowledge domains showing lower accuracy (53.2% vs. 62.2%, OR = 0.69, p = 0.05). A high consistency rate of 91.6% in explanations across all domains indicates reliable delivery of coherent and logical information. The simple prompt outperformed strategy-based prompts in accuracy (60% vs. 40%, p = 0.016), highlighting ChatGPT's limitations in its inability to accurately self-assess uncertainty and a tendency towards overconfidence, which may hinder medical decision-making. ChatGPT-4's high accuracy and consistent explanations in urology board examination demonstrate its potential in medical information processing. However, its limitations in self-assessment and overconfidence necessitate caution in its application, especially for inexperienced users. These insights call for ongoing advancements of urology-specific AI tools.

Experience of Data Analysis application for proceeding research data in solving the problem of target age group establishment

The article presents the experience of artificial intelligence application in research process. The article contains general information about basic concepts of machine learning (clustering and visualization), as well as considers more detaily an experience of clinical testing. The effectiveness of applying Data Analysis methods and means as one of the research stages is demonstrated on the example of a case on processing medical information using algorithms of machine learning: solving the problem on diagnostic value of the proposed indicator (FTF) for determining target age groups. Implementation of such approach of digital transformation improves the operational effectiveness of researches, as well as quality and availability of final technological products being developed - software for solving expert problems.

Developing and Analyzing Deep Learning and Natural Language Processing Systems in the Context of Medical Information Processing

This work aims to develop and analyze deep learning and natural language processing systems in the context of medical information processing. The amount of data created about patients in the healthcare system is always increasing. The human review of this enormous volume of data derived from numerous sources is expensive and takes a lot of time. Additionally, during a patient visit, doctors write down the patient’s medical encounter and send it to nurses and other medical departments for processing. Often, the doctor doesn’t have enough time to record every observation made while examining the patient and asking about their medical history which takes time for a medical diagnosis to be made. The manual review of this vast amount of data generated from multiple sources is costly and very time-consuming. It brings huge challenges while attempting to review this data meaningfully. Therefore, the goal of this research is to create a system that will address the aforementioned issues. The suggested method extracts voice data from medical encounters and converts it to text using Deep Learning (DL) and Natural Language Processing (NLP) techniques. More so, the system developed will improve medical intelligence processing by using deep learning to analyze medical datasets and produce results of a diagnosis, assisting medical professionals at various levels in making realistic, intelligent decisions in real-time regarding crucial health issues. The system was designed using the Object-Oriented Analysis and Design Methodology (OOADM), and the user interfaces were put into place utilizing Natural Language Processing techniques, particularly speech recognition and natural language comprehension. Speech recognition allows for the taking of free text notes, which can drastically cut down on the amount of time medical staff spends on labor-in the tensive clinical recording. By extracting different pieces of data for medical diagnosis and producing results in a matter of seconds, a deep learning algorithm demonstrates a significant capacity to construct clinical decision support systems. The system’s results demonstrate that the deep learning algorithm enabled medical intelligence to be 96.7 percent accurate.

ІНФОРМАЦІЙНА ТЕХНОЛОГІЯ ОПРАЦЮВАННЯ МЕДИЧНИХ ПОКАЗНИКІВ

The article presents a new information technology for processing medical indicators, which is based on the method of his-togram analysis. Histogram analysis, known as a key tool in medical statistics, allows visually and quantitatively evaluate the distribution of data, which becomes important for detecting asymmetry, deviations and other characteristics of the indicator base.The analysis of the existing algorithms for presenting and processing medical information shows that they do not fully satisfy the requirements for solving problems that require complex logical conclusions, taking into account the incompleteness and in-consistencies of the input data. To solve this problem, it is necessary to look for new technologies for processing medical data.It is proposed to use a histogram analysis of each indicator for the studied groups: a group of healthy patients (M1), a group of sick patients who received DASH treatment (M2), and a group of sick patients who were treated by the standard method (M3).The purpose of the article is to develop an information technology for the analysis of the probability of changes in the values of medical indicators in the course of the chosen method of treatment.The mathematical justification and algorithm of information technology for calculating medical indicators using the histogram method have been developed, fragmentsof normalized histograms and the obtained results are given. Information technology for processing medical indicators based on the method of histogram analysis is presented.Histogram analysis is an important tool in medical statistics and research. This method allows visually and quantitatively evaluate the distribution of data, which is useful for detecting deviations, asymmetry and other characteristics of distributions.Analysis of the indicators obtained by two different treatment methods showed both positive and negative changes, which require further research. A feature of the proposed information technology is that it allows assessing the state of the indicator as a whole, without reference to a specific patient

TeaBERT: An Efficient Knowledge Infused Cross-Lingual Language Model for Mapping Chinese Medical Entities to the Unified Medical Language System.

Medical entity normalization is an important task for medical information processing. The Unified Medical Language System (UMLS), a well-developed medical terminology system, is crucial for medical entity normalization. However, the UMLS primarily consists of English medical terms. For languages other than English, such as Chinese, a significant challenge for normalizing medical entities is the lack of robust terminology systems. To address this issue, we propose a translation-enhancing training strategy that incorporates the translation and synonym knowledge of the UMLS into a language model using the contrastive learning approach. In this work, we proposed a cross-lingual pre-trained language model called TeaBERT, which can align synonymous Chinese and English medical entities across languages at the concept level. As the evaluation results showed, the TeaBERT language model outperformed previous cross-lingual language models with Acc@5 values of 92.54%, 87.14% and 84.77% on the ICD10-CN, CHPO and RealWorld-v2 datasets, respectively. It also achieved a new state-of-the-art cross-lingual entity mapping performance without fine-tuning. The translation-enhancing strategy is applicable to other languages that face the similar challenge due to the absence of well-developed medical terminology systems.

Future Perspectives: A Literature Review of the Urologist’s New Potential Tool-Artificial Intelligence

Introduction: The overview aims to present the current state of artificial intelligence (AI) tools in the decision-making process, diagnosis, treatment possibilities, or outcome predictions in functional urology. Terminology: Artificial intelligence in healthcare includes computer systems that use reasoning and learning to analyze complex medical data and make predictions. Machine learning focuses on training computers with data to improve performance over time, especially in the case of processing medical information. Deep learning uses complex neural networks, inspired by the functional structure of the human brain, to extract information from medical data, improving the accuracy of diagnoses and speeding up investigations. Current AI Applications in Urology: The integration of artificial intelligence (AI) and artificial neural networks (ANN) in urological practice as a branch of medicine represents a promising potential for improving various aspects of patient care, diagnosis, surgical approaches with subsequent procedures resulting from rational and quick decision-making in the management of urological diseases (urolithiasis, BPH, oncology). AI technologies provide the opportunity to analyze large volumes of complex data, assist in clinical decision-making, optimize treatment strategies, and improve patient outcomes. Conclusion: The future and potential prospects of artificial intelligence as an auxiliary tool are significant, and the complete transition to them is still ahead. The field of AI support in urology is rapidly evolving, with continuous research and development focused on increasing the accuracy of diagnostic tools, refining surgical techniques, and improving treatment optimization. Continued research and clinical trials will be crucial for validating the effectiveness of AI and directing their integration into everyday urological practice.

Efficient Low-Resource Medical Information Processing Based on Semantic Analysis and Granular Computing

The rise of the digital economy and e-commerce has fostered a movement towards efficient low-resource medical information processing, a trend that holds great importance in the healthcare sector. Diabetes, being a widespread chronic condition, has witnessed the introduction of glucometers, which offer patients a convenient method of monitoring their blood sugar levels. However, it is worth noting that a considerable proportion of online comments may be subject to emotional bias or contain inaccurate information. Furthermore, the performance of glucometers can be influenced by several attributes, including price, accuracy and portability, thereby potentially complicating the decision-making process for consumers. Semantic analysis can be employed to acquire valuable information, aiding consumers in reasonably choosing the suitable glucometer. This paper utilizes the benefits of granular computing, an emerging computing paradigm, to effectively handle incomplete and uncertain medical information. It employs generalized fuzzy sets, rough sets and three-way decisions (TWD) techniques to boost the accuracy and reliability of medical information fusion. Subsequently, the MABAC (Multi-Attribute Border Approximation Area Comparison) method is utilized to evaluate the reviews of every glucometer, calculate their aggregated scores, and rank and compare them. Ultimately, in light of consumers’ needs and trade-offs, the glucometer with the highest score can be selected. The proposed approach comprehensively considers the weight and priority of multiple attributes, reduces information overload and mitigates selection difficulties, thereby enhancing the accuracy and reliability of low-resource medical information processing.

Digital Health Information Systems in the Member States of the Commonwealth of Independent States: Status and Prospects

This paper examines the status of the development of national digital health information systems (HIS) in Commonwealth of Independent States (CIS) member states. Data for research were collected using a questionnaire adapted from the questionnaire of the WHO’s Third Global Survey on eHealth. The results showed that the digital transformation of HIS was occurring in all seven CIS member states (participating in the study), which were financed by different resources. Laws and regulations on electronic medical records (EMR) use were present in almost all participating CIS member states. Various international standards and classifications were used to support development and the interoperability of digital health information system (d-HIS), including International Classification of Diseases (ICD), Digital Imaging and Communications in Medicine (DICOM), ISO 18308, Logical Observation Identifiers, Names, and Codes (LOINC), Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT), and ISO TC 215. Several CIS member states had adopted a national information security strategy for the safe processing of both personal data and medical confidential information. The digital transformation of healthcare and the Empowerment through Digital Health initiative are taking place in all CIS member states, which are at different stages of introducing electronic medical and health records.

“Why don't they talk to our daughter?”: Eye-tracking AAC and medical communication in Rett syndrome

This investigation explores medical advocacy for Rett syndrome—a rare neurological disorder causing loss of speech and hand function—through the lens of eye-tracking augmentative and alternative communication (AAC) technology. We gathered data from five individuals with Rett syndrome and their caregivers, encompassing semi-structured virtual interviews, video analyses of AAC device usage, and examination of AAC pageset screenshots. The findings reveal diverse AAC strategies employed to express discomfort or illness and highlight caregivers' pivotal role in processing medical information. Notwishstanding challenges like cost, time limitations in medical contexts, training needs, and the lack of standardized AAC symptom descriptions, eye-tracking AAC technology has the potential to enhance symptom assessment, foster patient autonomy, and facilitate personalized medical care. This study illuminates the transformative power of this AAC technology in medical communication, showcasing its promise in tackling communication challenges and underscoring its capacity to enhance quality of life for those with Rett syndrome and similar health conditions.

Free-Space Optical Communication with an Optimized Lipschitz Exponent for Biosignal Telemetry

Abstract Healthcare monitoring is a rapidly developing network in the field of advanced medical treatment. The network combines the ideology of wireless communication, signal processing, medical information and real-time processing units to support the medical monitoring system. The proposed work focuses on the development of a Free-Space Optical (FSO) system to transmit the biosignals from a remote distance to the physician. Generally, the data transmitted over the FSO system is affected by various atmospheric conditions such as air medium, O2, and H2O molecules. To tackle these problems, the Biosignals Electrocardiogram (ECG) and Electroencephalogram (EEG) are processed in the Optimized Lipschitz Exponent (OLE) function before transmission over the FSO medium. In this novel technique, the OLE function measures the informative data from the biosignals by calculating the local regularities and singularity. This collects the most informative signals and transmits them in the signal over the FSO medium. This particular hybridization helps to transmit the required data without distortion. The Bit Error Rate (BER) of 10−9 is obtained, which satisfies the healthcare monitoring condition. The result section shows that the proposed model has minimum losses compared to the original signal.

Understanding of Machine Learning with Deep Learning: Architectures, Workflow, Applications and Future Directions

In recent years, deep learning (DL) has been the most popular computational approach in the field of machine learning (ML), achieving exceptional results on a variety of complex cognitive tasks, matching or even surpassing human performance. Deep learning technology, which grew out of artificial neural networks (ANN), has become a big deal in computing because it can learn from data. The ability to learn enormous volumes of data is one of the benefits of deep learning. In the past few years, the field of deep learning has grown quickly, and it has been used successfully in a wide range of traditional fields. In numerous disciplines, including cybersecurity, natural language processing, bioinformatics, robotics and control, and medical information processing, deep learning has outperformed well-known machine learning approaches. In order to provide a more ideal starting point from which to create a comprehensive understanding of deep learning, also, this article aims to provide a more detailed overview of the most significant facets of deep learning, including the most current developments in the field. Moreover, this paper discusses the significance of deep learning and the various deep learning techniques and networks. Additionally, it provides an overview of real-world application areas where deep learning techniques can be utilised. We conclude by identifying possible characteristics for future generations of deep learning modelling and providing research suggestions. On the same hand, this article intends to provide a comprehensive overview of deep learning modelling that can serve as a resource for academics and industry people alike. Lastly, we provide additional issues and recommended solutions to assist researchers in comprehending the existing research gaps. Various approaches, deep learning architectures, strategies, and applications are discussed in this work.

Digitalization in health care: Promising tools

Aim. The presented study aims to consider the possibilities and features of the development of digitalization in health care under modern conditions (through the example of Russia).Tasks. The authors describe the essence and trends of digitalization and the industry-specific features of its manifestation in health care; analyze the available digital technologies from the perspective of their use in health care; identify directions for the development of RWD and RWE systems in Russia and assess the need for a transformation of the institutional environment of health care in the context of their introduction.Methods. This study uses general scientific research methods (analysis, synthesis, monographic method, grouping), as well as methods of institutional and sectoral techno-economic analysis.Results. The process of digital transformation in the health care sector is analyzed, and the introduction of RWD and RWE systems and artificial intelligence systems is examined in detail. In the context of a study of development trends in digital health care, recommendations for improving the collection, processing, and storage of medical information are given. New means of working with medical data and directions for the development of digital medical services and medical care using digital technologies are described.Conclusions. Digitalization in health care is aimed at improving the efficiency and effectiveness of medical care and medical services through the application of innovative approaches and the expanded use of digital tools. Global experience shows that patients and medical organizations can successfully use personalized solutions in the field of digital health care to improve disease prevention, self-care, and access to health care, providing that they are designed with allowance for the needs and preferences of the patient. However, these issues require improved processes not only at the level of organizations, but also at the level of industrial regulatory bodies.

Sharing and Cooperation of Improved Cross-Entropy Optimization Algorithm in Telemedicine Multimedia Information Processing.

In order to improve the efficiency of medical multimedia information sharing, this paper combines cloud computing technology and SOA (service-oriented architecture) technology to build a medical multimedia information sharing system. Building a medical information sharing platform requires integrating information resources stored in information systems of medical institutions and nonmedical information systems related to medical information and forming a huge resource pool. It is important to mine and analyze the information resources in the resource pool to realize the sharing and interaction of medical information. To this end, this paper proposes a gain-adaptive control algorithm with online adjustable parameters and investigates the extension of the mutual entropy optimization algorithm in the control domain and its integrated processing capability in the process of medical multimedia information processing. In addition, this paper constructs a medical multimedia information sharing and collaboration platform with medical multimedia information sharing and telemedicine as the core and verifies the effectiveness of the platform through experiments. The simulation results and comparison results with other systems prove that the system in this paper can realize fast data processing, retrieve and analyze massive data, and meet the demand of remote intelligent diagnosis under the premise of safety and stability. Meanwhile, the system in this paper can help hospitals achieve fast and accurate diagnosis, which has strong theoretical and practical values.