Medical Information Retrieval Research Articles

Privacy-preserving large language models for structured medical information retrieval

Most clinical information is encoded as free text, not accessible for quantitative analysis. This study presents an open-source pipeline using the local large language model (LLM) “Llama 2” to extract quantitative information from clinical text and evaluates its performance in identifying features of decompensated liver cirrhosis. The LLM identified five key clinical features in a zero- and one-shot manner from 500 patient medical histories in the MIMIC IV dataset. We compared LLMs of three sizes and various prompt engineering approaches, with predictions compared against ground truth from three blinded medical experts. Our pipeline achieved high accuracy, detecting liver cirrhosis with 100% sensitivity and 96% specificity. High sensitivities and specificities were also yielded for detecting ascites (95%, 95%), confusion (76%, 94%), abdominal pain (84%, 97%), and shortness of breath (87%, 97%) using the 70 billion parameter model, which outperformed smaller versions. Our study successfully demonstrates the capability of locally deployed LLMs to extract clinical information from free text with low hardware requirements.

PubMed Retrieval with RAG Techniques.

This study explores the application of Retriever-Augmented Generation (RAG) in enhancing medical information retrieval from the PubMed database. By integrating RAG with Large Language Models (LLMs), we aim to improve the accuracy and relevance of medical information provided to healthcare professionals. Our evaluation on a labeled dataset of 1,000 queries demonstrates promising results in answer relevance, while highlighting areas for improvement in groundedness and context relevance.

Understanding Medical Professionals’ Views on Integrating Artificial Intelligence (AI) in Medical Information Retrieval and Healthcare Information Systems

Understanding Medical Professionals’ Views on Integrating Artificial Intelligence (AI) in Medical Information Retrieval and Healthcare Information Systems

A Descriptive Study to Assess the Attitude and Skill in Usage of Computer Among Registered Nurses in BMCH General Ward, Chennai

In recent years, the use of computers in the healthcare system has become more important for easy retrieval of medical information. This advancement has a positive impact on patient care among health care providers. Objectives: To assess the level of attitude and skill in use of computer among the nurses, to find the correlation between attitude and skill in use of computer among the nurses, to find out the association between the level of attitude and skill in use of computers with the selected demographic variables. Methodology: An Exploratory Descriptive method was used and the study was conducted in General wards at BMCH, Chennai. 110 registered nurses were selected using convenient sampling technique. 3-point Likert's scale was used to assess the level of attitude and observational checklist for assessing the level of skill in use of computers among registered nurses. Results: The study revealed 68 (62%) had favourable attitude, 31(28%) had most favourable attitude and 11(10%) had unfavourable attitude, 53(48%) had adequate skill, 38(35%) had moderately adequate skill, 19 (17%) had inadequate skill in use of computer. There was a correlation between level of attitude and skill in use of computer at the significant level of p value 0.021. There was a significant association between age and educational qualification with level of attitude and there was a significant association between age, area of work and attended training programme. Conclusion: The study result showed that 62% of registered nurses favours using computers for patient care, 48% possessed adequate skills which revealed a gap of 14%. The study suggested to conduct in-service education and training program to promote the use of computers in patient care. Keywords: Use of computer, Attitude, Skill.

A Liver Cancer Question-Answering System Based on Next-Generation Intelligence and the Large Model Med-PaLM 2

Chronic diseases are the "number one killer" threatening human life and health. In recent years, the incidence of chronic diseases has been rising, and the trend is younger, and the prevention and control situation is very serious. This study introduces a Liver Cancer Question-Answering System (LCQAS) that leverages next-generation artificial intelligence and the large model Med-PaLM 2 The Med-PaLM 2 medical question answering system, powered by next-generation intelligence and large language models, represents a cutting-edge tool in the healthcare domain. Leveraging advanced artificial intelligence techniques and the capabilities of large language models like Med-PaLM 2, this system aims to provide accurate and comprehensive responses to medical inquiries and queries. With its intuitive interface and contextually relevant answers, LCQAS represents a significant advancement in medical information retrieval for liver cancer management.

Leveraging Large Language Models for Clinical Abbreviation Disambiguation.

Clinical abbreviation disambiguation is a crucial task in the biomedical domain, as the accurate identification of the intended meanings or expansions of abbreviations in clinical texts is vital for medical information retrieval and analysis. Existing approaches have shown promising results, but challenges such as limited instances and ambiguous interpretations persist. In this paper, we propose an approach to address these challenges and enhance the performance of clinical abbreviation disambiguation. Our objective is to leverage the power of Large Language Models (LLMs) and employ a Generative Model (GM) to augment the dataset with contextually relevant instances, enabling more accurate disambiguation across diverse clinical contexts. We integrate the contextual understanding of LLMs, represented by BlueBERT and Transformers, with data augmentation using a Generative Model, called Biomedical Generative Pre-trained Transformer (BIOGPT), that is pretrained on an extensive corpus of biomedical literature to capture the intricacies of medical terminology and context. By providing the BIOGPT with relevant medical terms and sense information, we generate diverse instances of clinical text that accurately represent the intended meanings of abbreviations. We evaluate our approach on the widely recognized CASI dataset, carefully partitioned into training, validation, and test sets. The incorporation of data augmentation with the GM improves the model's performance, particularly for senses with limited instances, effectively addressing dataset imbalance and challenges posed by similar concepts. The results demonstrate the efficacy of our proposed method, showcasing the significance of LLMs and generative techniques in clinical abbreviation disambiguation. Our model achieves a good accuracy on the test set, outperforming previous methods.

The application of NLP in information retrieval

The field of Natural Language Processing (NLP) has experienced impressive advancements and has found diverse applications. This paper presents a comprehensive review of the development of NLP in the field of information retrieval. It explores different stages of NLP techniques and methods, including keyword matching, rule-based approaches, statistical methods, and the utilization of machine learning and deep learning technologies. Furthermore, the paper provides detailed insights into the specific applications of NLP in domains such as academic information retrieval, medical information retrieval, travel information retrieval, and e-commerce information retrieval. It analyzes the current state of NLP applications in these domains, highlights their advantages, and discusses their associated limitations. Finally, the paper emphasizes the continuous advancement of the NLP field, with a particular focus on semantic understanding, personalized retrieval, and multimodal information retrieval, to better adapt to diverse data types and user requirements. The paper concludes by summarizing the main points discussed and providing future directions.

(021) ChatGPT's Ability to Assess Quality and Readability of Online Medical Information

Abstract Introduction Health literacy plays a crucial role in enabling patients to understand and effectively use medical information. As technology rapidly advances, the significance of health literacy becomes even more pronounced, particularly in comprehending complex medical information. Artificial Intelligence (AI) platforms have garnered significant attention for their remarkable ability to generate automated responses to a wide range of prompts. However, their capacity to assess the quality and readability of provided text remains uncertain. Given the growing prominence of AI web assistant tools, we hypothesized that integrating these tools into patients' web searches could enhance the retrieval of accurate medical information. Objective To evaluate the proficiency of Conversational Generative Pre-Trained Transformer (ChatGPT) in assessing readability, and utilizing the DISCERN tool to assess quality of online content regarding shock wave therapy for erectile dysfunction. Methods Websites were generated using a Google search of “shock wave therapy for erectile dysfunction” with location filters disabled. Readability was analyzed using the Readable software (Readable.com, Horsham, United Kingdom). Quality was assessed independently by three reviewers using the DISCERN tool. The same plain text files collected were inputted into ChatGPT to determine whether it produced comparable metrics for readability and quality. Results The study results revealed a notable disparity between ChatGPT's readability assessment and that obtained from a reliable tool, Readable.com (p<0.05). This indicates a lack of alignment between ChatGPT's algorithm and that of established tools, such as Readable.com. Similarly, the DISCERN score generated by ChatGPT differed significantly from the scores generated manually by human evaluators (p<0.05), suggesting that ChatGPT may not be capable of accurately identifying poor-quality information sources regarding shock wave therapy as a treatment for erectile dysfunction. Conclusions ChatGPT’s evaluation of the quality and readability of online text regarding shockwave therapy for erectile dysfunction differs from that of human raters and trusted tools. ChatGPT's current capabilities were not sufficient for reliably assessing the quality and readability of textual content. Further research is needed to elucidate the role of AI in the objective evaluation of online medical content in other fields. Continued development in AI and incorporation of tools such as DISCERN into AI software may enhance the way patients navigate the web in search of high-quality medical content in the future. Disclosure No.

Text clustering based on pre-trained models and autoencoders.

Text clustering is the task of grouping text data based on similarity, and it holds particular importance in the medical field. sIn healthcare, medical data clustering is a highly active and effective research area. It not only provides strong support for making correct medical decisions from medical datasets but also aids in patient record management and medical information retrieval. With the development of the healthcare industry, a large amount of medical data is being generated, and traditional medical data clustering faces significant challenges. Many existing text clustering algorithms are primarily based on the bag-of-words model, which has issues such as high dimensionality, sparsity, and the neglect of word positions and context. Pre-trained models are a deep learning-based approach that treats text as a sequence to accurately capture word positions and context information. Moreover, compared to traditional K-means and fuzzy C-means clustering models, deep learning-based clustering algorithms are better at handling high-dimensional, complex, and nonlinear data. In particular, clustering algorithms based on autoencoders can learn data representations and clustering information, effectively reducing noise interference and errors during the clustering process. This paper combines pre-trained language models with deep embedding clustering models. Experimental results demonstrate that our model performs exceptionally well on four public datasets, outperforming most existing text clustering algorithms, and can be applied to medical data clustering.

SISTEM ANALISIS PENYAKIT MATA BERBASIS PHP DAN MYSQL MENGGUNAKAN METODE TF-IDF

The development of information technology has had a positive impact on the healthcare sector, particularly in disease analysis. Eye diseases are among the health issues that require early detection for effective prevention and treatment. In this context, this research proposes the development of a PHP and MySQL-based Eye Disease Analysis System utilizing the Term Frequency-Inverse Document Frequency (TF-IDF) method. The system is designed to provide a web platform facilitating the analysis of textual data related to eye diseases. Involving PHP as the programming language and MySQL as the database, the system allows efficient storage, management, and retrieval of medical information. The TF-IDF method is employed to analyze unique characteristics within textual data, enabling the identification of symptoms, diagnosis, and treatment of eye diseases. Context Diagram, Data Flow Diagram (DFD), and Entity-Relationship Diagram (ERD) serve as visual guides in designing this system. The Context Diagram provides a general overview of the system's interaction with its external environment, while DFD illustrates the flow of data within the system. ERD is used to represent the data structure within the database.
 The research results are expected to significantly contribute to the efficiency of eye disease analysis, providing faster information to medical professionals and enhancing the general public's understanding of eye diseases. This system is anticipated to serve as a foundation for further developments in the application of information technology in the healthcare sector, particularly in disease analysis.

Assessing the performance of ChatGPT's responses to questions related to epilepsy: A cross-sectional study on natural language processing and medical information retrieval

BackgroundEpilepsy is a neurological condition marked by frequent seizures and various cognitive and psychological effects. Reliable information is essential for effective treatment. Natural language processing models like ChatGPT are increasingly used in healthcare for information access and data analysis, making it crucial to assess their accuracy. ObjectiveThis study aimed to investigate the accuracy of ChatGPT in providing educational information related to epilepsy. MethodsWe compared the answers from ChatGPT-4 and ChatGPT-3.5 to 57 common epilepsy questions based on the Korean Epilepsy Society's "Epilepsy Patient and Caregiver Guide." Two epileptologists reviewed the responses, with a third serving as an arbiter in cases of disagreement. ResultsOut of 57 questions, 40 responses from ChatGPT-4 had "sufficient educational value," 16 were "correct but inadequate," and one was "mixed with correct and incorrect" information. No answers were entirely incorrect. GPT-4 generally outperformed GPT-3.5 and was often on par with or better than the official guide. ConclusionsChatGPT-4 shows promise as a tool for delivering reliable epilepsy-related information and could help alleviate the educational burden on healthcare professionals. Further research is needed to explore the benefits and limitations of using such models in medical contexts.

Medical Information Retrieval with Weighting Critical Score for Acute Respiratory Infection (ARI) Desease Detection

Medical Information Retrieval (Med-IR )is part of computer science that discusses the search for a medical document. Medical Information Retrieval is needed by patients to know the initial prediction of the symptoms they are experiencing. ARI (Acute Respiratory Infection) is a disease that almost everyone has experienced which can cause death. This study uses a dataset of ARI sufferers and user queries that contain symptoms in text form. Furthermore, the query data is processed with the Med-IR application using Bi-Gram, TF-IDF as the feature extraction and Cosine Similarity as the similarity method, so that a return document is produced which is expected to be used as an early prediction of ARI in patients. The research also uses a critical disease wighting process, so that the results of the Med-IR are complemented by predictions of the severity level of the disease. From the results of research conducted at the Assyafi'u Sentosa Lengkong Clinic, Nganjuk, the best results were obtained for precision values ​​of 85.5% and 52.9% for recall values ​. The evaluation of disease severity with Mean Absolute Percentage Error (MAPE) getting a low score of 2,529%. Keyword : Medical Information Retrieval, ARI, Weighting critical disease, Bi-Gram, TF-IDF, Cosine Similarity.

Consultants’ and referrers’ perceived barriers to closing the cross-institutional referral loop, and perceived impact on clinical care

BackgroundCross-institutional (external) referrals are prone to communication breakdowns, increasing patient safety risks, clinician burnout, and healthcare costs. To close these external referral loops, referring primary care physicians (PCPs) need to receive patient information from consultants at different healthcare institutions. Although existing studies investigated the early phases of external referral loops, we lack sufficient knowledge about the closing phases of these loops. This knowledge could allow health care institutions to improve care coordination and rates of closed referral loops by implementing socio-technical interventions for patient information exchange throughout a referral loop. Human factors engineering (HFE) provides a systematic approach to advance our understanding of barriers perceived by physicians. Using HFE, our objective was to characterize referring and consulting physicians’ barriers to closing referral loops and implications for care. MethodsThis qualitative cross-sectional study included semi-structured interviews with referrers and external consultants. We used the Systems Engineering Initiative for Patient Safety 2.0 framework to conduct rapid qualitative analyses, determining perceived barriers and related implications. Main measures were consultants’ and referrers’ perceptions of, and experiences with, barriers to external referrals. ResultsSix referring PCPs and 12 consultants participated from two healthcare systems and four medical specialties. Physicians perceived three main barriers in external referrals: receipt of excessive and unnecessary faxed documents, missing or delayed documentation, and organizational policies regarding information privacy interfering with closing the loop. Compared to internal referrals, physicians reported increased staff burden, patient frustration, and delays in diagnosis with external referrals. Consultants reported the ability to provide the same level of care to patients with internal or external referrals. However, consultants described communication breakdowns that prohibited confirmation of follow-up plan retrieval, initiation, or effectiveness. ConclusionPhysicians reported technological and organizational barriers to closing cross-institutional referral loops. Promises of HIE technology for external referrals have not fully materialized. Among physicians and patients, retrieval and exchange of medical information increases perceived workload, burden, and frustration. These increases are not accurately captured by traditional organizational metrics. This study provides evidence that informs future human factors engineering research to address perceived barriers and guide future HIE design or implementation.

Fog-Assisted Internet of Medical Things for Smart Healthcare

The Internet of Medical Things (IoMT) is used to sense and deliver vital signs for healthcare monitoring, so rapid medical information retrieval is crucial. The information-centric model can help shorten delays of IoMT-based medical data delivery, but IoMT nodes cannot perform the information-centric model due to their limited resources. In this paper, we integrate IoMT with fog clouds and propose a fog-assisted information-centric IoMT framework so that fog nodes can help perform the information-centric model and reduce latency of IoMT-based medical information retrieval. In the framework, consumers can exploit reverse-paths and request aggregation to acquire medical information of multiple subjects from the nearest provider via one information delivery procedure. Moreover, the framework addresses the consumer mobility issue to guarantee receipt of medical information. Finally, this framework is evaluated to verify its advantages.

Attitudes and Perceptions of Postgraduate Medical Trainees towards the Library User Education (LUE) Programme

Library user education aims to make library users proficient in research and lifelong learners. The impact of library user education extends beyond the classroom and orientation, allowing users to engage themselves and effectively identify, access, retrieve, and use library information resources to their advantage. It also raises awareness of the library's available information resources.The Postgraduate Institute of Medicine (PGIM) library at the University of Colombo has identified the importance of acquiring user perceptions through user education programmes. This study aimed to assess postgraduate trainees' perceptions and investigate the impact of the user education programme conducted by the PGIM library. The library user education programmes considered in this study are the PGIM library resources and services, which include the library digital repository, literature review, E-Resources, avoid in gplagiarism and managing references. The survey research design was used for the study, while the questionnaire instrument was used for data collection.The study population comprises all users participated for the library user education programme in 2020 -2021. The study found out that the user education programme had positively affected the postgraduate trainees' use of the library and academic performance. Awareness, accessibility, and utilisation of library services and resources impact the library user education programme.The study also emphasised the significance of ongoing library user education programmes in developing research skills among medical students in order for them to develop medical information retrieval, supporting research work and other findings provided to medical trainees, which supports their professional life.

Effect of Artificial Intelligence-based Health Education Accurately Linking System (AI-HEALS) for Type 2 diabetes self-management: protocol for a mixed-methods study

BackgroundPatients with type 2 diabetes (T2DM) have an increasing need for personalized and Precise management as medical technology advances. Artificial intelligence (AI) technologies on mobile devices are being developed gradually in a variety of healthcare fields. As an AI field, knowledge graph (KG) is being developed to extract and store structured knowledge from massive data sets. It has great prospects for T2DM medical information retrieval, clinical decision-making, and individual intelligent question and answering (QA), but has yet to be thoroughly researched in T2DM intervention. Therefore, we designed an artificial intelligence-based health education accurately linking system (AI-HEALS) to evaluate if the AI-HEALS-based intervention could help patients with T2DM improve their self-management abilities and blood glucose control in primary healthcare.MethodsThis is a nested mixed-method study that includes a community-based cluster-randomized control trial and personal in-depth interviews. Individuals with T2DM between the ages of 18 and 75 will be recruited from 40-45 community health centers in Beijing, China. Participants will either receive standard diabetes primary care (SDPC) (control, 3 months) or SDPC plus AI-HEALS online health education program (intervention, 3 months). The AI-HEALS runs in the WeChat service platform, which includes a KBQA, a system of physiological indicators and lifestyle recording and monitoring, medication and blood glucose monitoring reminders, and automated, personalized message sending. Data on sociodemography, medical examination, blood glucose, and self-management behavior will be collected at baseline, as well as 1,3,6,12, and 18 months later. The primary outcome is to reduce HbA1c levels. Secondary outcomes include changes in self-management behavior, social cognition, psychology, T2DM skills, and health literacy. Furthermore, the cost-effectiveness of the AI-HEALS-based intervention will be evaluated.DiscussionKBQA system is an innovative and cost-effective technology for health education and promotion for T2DM patients, but it is not yet widely used in the T2DM interventions. This trial will provide evidence on the efficacy of AI and mHealth-based personalized interventions in primary care for improving T2DM outcomes and self-management behaviors.Trial registrationBiomedical Ethics Committee of Peking University: IRB00001052-22,058, 2022/06/06; Clinical Trials: ChiCTR2300068952, 02/03/2023.

IoT-Enabled Classification of Echocardiogram Images for Cardiovascular Disease Risk Prediction with Pre-Trained Recurrent Convolutional Neural Networks

Cardiovascular diseases currently present a key health concern, contributing to an increase in death rates worldwide. In this phase of increasing mortality rates, healthcare represents a major field of research, and the knowledge acquired from this analysis of health information will assist in the early identification of disease. The retrieval of medical information is becoming increasingly important to make an early diagnosis and provide timely treatment. Medical image segmentation and classification is an emerging field of research in medical image processing. In this research, the data collected from an Internet of Things (IoT)-based device, the health records of patients, and echocardiogram images are considered. The images are pre-processed and segmented, and then further processed using deep learning techniques for classification as well as forecasting the risk of heart disease. Segmentation is attained via fuzzy C-means clustering (FCM) and classification using a pretrained recurrent neural network (PRCNN). Based on the findings, the proposed approach achieves 99.5% accuracy, which is higher than the current state-of-the-art techniques.

Nature-inspired query optimisation models for medical information retrieval with relevance feedback

Nature-inspired query optimisation models for medical information retrieval with relevance feedback

Nature-inspired query optimisation models for medical information retrieval with relevance feedback

Nature-inspired query optimisation models for medical information retrieval with relevance feedback

Drug Adverse Event Detection Using Text-Based Convolutional Neural Networks (TextCNN) Technique

With the rapid advancement in healthcare, there has been exponential growth in the healthcare records stored in large databases to help researchers, clinicians, and medical practitioner’s for optimal patient care, research, and trials. Since these studies and records are lengthy and time consuming for clinicians and medical practitioners, there is a demand for new, fast, and intelligent medical information retrieval methods. The present study is a part of the project which aims to design an intelligent medical information retrieval and summarization system. The whole system comprises three main modules, namely adverse drug event classification (ADEC), medical named entity recognition (MNER), and multi-model text summarization (MMTS). In the current study, we are presenting the design of the ADEC module for classification tasks, where basic machine learning (ML) and deep learning (DL) techniques, such as logistic regression (LR), decision tree (DT), and text-based convolutional neural network (TextCNN) are employed. In order to perform the extraction of features from the text data, TF-IDF and Word2Vec models are employed. To achieve the best performance of the overall system for efficient information retrieval and summarization, an ensemble strategy is employed, where predictions of the selected base models are integrated to boost the robustness of one model. The performance results of all the models are recorded as promising. TextCNN, with an accuracy of 89%, performs better than the conventional machine learning approaches, i.e., LR and DT with accuracies of 85% and 77%, respectively. Furthermore, the proposed TextCNN outperforms the existing adverse drug event classification approaches, achieving precision, recall, and an F1 score of 87%, 91%, and 89%, respectively.