International Journal of Medical Informatics最新文献

{"title":"PlasmaCell CAD: A computer-aided diagnosis software tool for plasma cell recognition and characterization in microscopic images","authors":"Rasoul Kasbgar ,&nbsp;Alireza Vard","doi":"10.1016/j.ijmedinf.2025.105869","DOIUrl":"10.1016/j.ijmedinf.2025.105869","url":null,"abstract":"<div><h3>Background and objective</h3><div>In the traditional diagnostic process for multiple myeloma cancer, a pathologist screens prepared blood samples using a microscope to detect, classify, and count plasma cells. This manual approach is time-consuming, exhausting, and prone to human errors. Consequently, medical experts and researchers are highly interested in any tool that partially or entirely automates this process. To achieve this goal, we developed a software tool called PlasmaCell CAD to analyze effective cells for diagnosing multiple myeloma cancers through microscopic images.</div></div><div><h3>Methods</h3><div>In the proposed software, to detect and segment cells, we exploit the Mask-RCNN model that has been enhanced by leveraging the circlet transform for the anchor generation. Also, we use the SVM classifier to identify normal and abnormal plasma cells in this software. Moreover, we designed and developed a graphical user interface (GUI) for the PlasmaCell CAD so that users would be able to work with it more easily.</div></div><div><h3>Results</h3><div>we considered the performance of the proposed software on both a publicly available dataset and a locally collected dataset. The experimental results demonstrated the capability and efficiency of PlasmaCell CAD software in segmenting and classifying plasma cells as well as its ease of use.</div></div><div><h3>Conclusions</h3><div>PlasmaCell CAD is a free software tool that can easily be downloaded and installed on any computers running Windows. PlasmaCell CAD provides a user-friendly GUI with several image processing and visualization facilities for the user that can accelerate the diagnosis process. In light of promising results, PlasmaCell CAD software can be useful to pathologists in helping to diagnose multiple myeloma cancer.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"198 ","pages":"Article 105869"},"PeriodicalIF":3.7,"publicationDate":"2025-03-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143580612","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Stress monitoring using low-cost electroencephalogram devices: A systematic literature review","authors":"Gideon Vos ,&nbsp;Maryam Ebrahimpour ,&nbsp;Liza van Eijk ,&nbsp;Zoltan Sarnyai ,&nbsp;Mostafa Rahimi Azghadi","doi":"10.1016/j.ijmedinf.2025.105859","DOIUrl":"10.1016/j.ijmedinf.2025.105859","url":null,"abstract":"&lt;div&gt;&lt;h3&gt;Introduction&lt;/h3&gt;&lt;div&gt;The use of low-cost, consumer-grade wearable health monitoring devices has become increasingly prevalent in mental health research, including stress studies. While cortisol response magnitude remains the gold standard for stress assessment, an expanding body of research employs low-cost EEG devices as primary tools for recording biomarker data, often combined with wrist and ring-based wearables. However, the technical variability among low-cost EEG devices, particularly in sensor count and placement according to the 10-20 Electrode Placement System, poses challenges for reproducibility in study outcomes.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Objective&lt;/h3&gt;&lt;div&gt;This review aims to provide an overview of the growing application of low-cost EEG devices and machine learning techniques for assessing brain function, with a focus on stress detection. It also highlights the strengths and weaknesses of various machine learning methods commonly used in stress research, and evaluates the reproducibility of reported findings along with sensor count and placement importance.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Methods&lt;/h3&gt;&lt;div&gt;A comprehensive review was conducted of published studies utilizing EEG devices for stress detection and their associated machine learning approaches. Searches were performed across databases including Scopus, Google Scholar, ScienceDirect, Nature, and PubMed, yielding 69 relevant articles for analysis. The selected studies were synthesized into four thematic categories: stress assessment using EEG, low-cost EEG devices, datasets for EEG-based stress measurement, and machine learning techniques for EEG-based stress analysis. For machine learning-focused studies, validation and reproducibility methods were critically assessed. Study quality was evaluated and scored using the IJMEDI checklist.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Results&lt;/h3&gt;&lt;div&gt;The review identified several studies employing low-cost EEG devices to monitor brain activity during stress and relaxation phases, with many reporting high predictive accuracy using various machine learning validation techniques. However, only 54% of the studies included health screening prior to experimentation, and 58% were categorized as low-powered due to limited sample sizes. Additionally, few studies validated their results using an independent validation set or cortisol response as a correlating biomarker and there was a lack of consensus on data pre-processing and sensor placement as a key contributor to improving model generalization and accuracy.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Conclusion&lt;/h3&gt;&lt;div&gt;Low-cost consumer-grade wearable devices, including EEG and wrist-based monitors, are increasingly utilized in stress-related research, offering promising avenues for non-invasive biomarker monitoring. However, significant gaps remain in standardizing EEG signal processing and sensor placement, both of which are critical for enhancing model generalization and accuracy. Furthermore, the limited use of independent validation ","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"198 ","pages":"Article 105859"},"PeriodicalIF":3.7,"publicationDate":"2025-03-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143580611","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Artificial intelligence based predictive tools for identifying type 2 diabetes patients at high risk of treatment Non-adherence: A systematic review","authors":"Malede Berihun Yismaw ,&nbsp;Chernet Tafere ,&nbsp;Bereket Bahiru Tefera ,&nbsp;Desalegn Getnet Demsie ,&nbsp;Kebede Feyisa ,&nbsp;Zenaw Debasu Addisu ,&nbsp;Tirsit Ketsela Zeleke ,&nbsp;Ebrahim Abdela Siraj ,&nbsp;Minichil Chanie Worku ,&nbsp;Fasikaw Berihun","doi":"10.1016/j.ijmedinf.2025.105858","DOIUrl":"10.1016/j.ijmedinf.2025.105858","url":null,"abstract":"<div><h3>Aims</h3><div>Several Artificial Intelligence (AI) based predictive tools have been developed to predict non-adherence among patients with type 2 diabetes (T2D). Hence, this study aimed to describe and evaluate the methodological quality of AI based predictive tools for identifying T2D patients at high risk of treatment non-adherence.</div></div><div><h3>Methods</h3><div>A systematic search was conducted across multiple databases including, EMBASE, Cochrane Library, MedLine, and Google Scholar search. The Prediction model Risk Of Bias ASsessment Tool (PROBAST) was used to assess the quality of studies. The performances of tools were assessed by Area Under the Curve (AUC), precision, recall, C-index, accuracy, sensitivity, specificity or F1 score.</div></div><div><h3>Results</h3><div>Most studies measured predictive ability using AUC (75 %), and some only reported precision (25 %), recall (12.5 %), C-index (12.5 %), accuracy (37.5), sensitivity (12.5 %), specificity (12.5 %) or F1 score (25 %). All tools had moderate to high predictive ability (AUC &gt; 0.70). However, only one study conducted external validation. Demographic characteristics, HbA1c, glucose monitoring data, and treatment details were typical factors used in developing tools.</div></div><div><h3>Conclusions</h3><div>The existing AI based tools holds significant promise for improving diabetes care. However, future studies should focus on refining the existing tools, validating in other settings, and evaluating the cost-effectiveness of AI-supported interventions.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"198 ","pages":"Article 105858"},"PeriodicalIF":3.7,"publicationDate":"2025-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143535390","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"RelAI: an automated approach to judge pointwise ML prediction reliability","authors":"Lorenzo Peracchio ,&nbsp;Giovanna Nicora ,&nbsp;Enea Parimbelli ,&nbsp;Tommaso Mario Buonocore ,&nbsp;Eleonora Tavazzi ,&nbsp;Roberto Bergamaschi ,&nbsp;Arianna Dagliati ,&nbsp;Riccardo Bellazzi","doi":"10.1016/j.ijmedinf.2025.105857","DOIUrl":"10.1016/j.ijmedinf.2025.105857","url":null,"abstract":"<div><h3>Objectives</h3><div>AI/ML advancements have been significant, yet their deployment in clinical practice faces logistical, regulatory, and trust-related challenges. To promote trust and informed use of ML predictions in real-world scenarios, reliable assessment of individual predictions is essential. We propose RelAI, a tool for pointwise reliability assessment of ML predictions that can support the identification of prediction errors during deployment.</div></div><div><h3>Materials and Methods</h3><div>RelAI utilizes Autoencoders (AEs) to detect distributional shifts (Density principle) and a proxy model to encode local performance (Local Fit principle). We validated RelAI on a synthetic dataset and a real-world scenario involving Multiple Sclerosis (MS) patient outcomes.</div></div><div><h3>Results</h3><div>On a synthetic dataset, RelAI effectively identified unreliable predictions, outperforming alternative approaches. In the MS case study, reliable predictions exhibited higher accuracy and were associated with specific demographic features, such as sex, residence, and eye symptoms.</div></div><div><h3>Discussion and Conclusion</h3><div>RelAI can support ML deployment in clinical settings by providing pointwise reliability assessments, ensuring regulatory compliance, and fostering user trust. Its model-agnostic nature and its compatibility with Python-based ML pipelines enhance its potential for widespread adoption.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"197 ","pages":"Article 105857"},"PeriodicalIF":3.7,"publicationDate":"2025-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143528721","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Hypothesis: Net benefit as an objective function during development of machine learning algorithms for medical applications","authors":"Andrew Vickers ,&nbsp;Alexander Hollingsworth ,&nbsp;Anthony Bozzo ,&nbsp;Avijit Chatterjee ,&nbsp;Subrata Chatterjee","doi":"10.1016/j.ijmedinf.2025.105844","DOIUrl":"10.1016/j.ijmedinf.2025.105844","url":null,"abstract":"<div><div>Net benefit is the most widely used metric for evaluating the clinical utility of medical prediction models. The approach applies decision analytic theory to weight true and false positives depending on the relative consequences of different decision outcomes. It is plausible that there are at least some machine learning scenarios where optimization of the objective function during model development will not optimize net benefit during model evaluation. We therefore hypothesize that optimizing net benefit during model development will in some cases ultimately lead to higher clinical utility than optimizing for mean square error or some other unweighted loss function. There is some preliminary evidence that this does indeed occur. We accordingly recommend further methodologic research to determine the use cases where net benefit should be the objective function during model development.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"197 ","pages":"Article 105844"},"PeriodicalIF":3.7,"publicationDate":"2025-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143508974","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing readmission prediction model in older stroke patients by integrating insight from readiness for hospital discharge: Prospective cohort study","authors":"Huixiu Hu ,&nbsp;Yajie Zhao ,&nbsp;Chao Sun ,&nbsp;Quanying Wu ,&nbsp;Ying Deng ,&nbsp;Jie Liu","doi":"10.1016/j.ijmedinf.2025.105845","DOIUrl":"10.1016/j.ijmedinf.2025.105845","url":null,"abstract":"<div><h3>Background</h3><div>The 30-day hospital readmission rate is a key indicator of healthcare quality and system efficiency. This study aimed to develop machine-learning (ML) models to predict unplanned 30-day readmissions in older patients with ischemic stroke (IS) using a prospective cohort design.</div></div><div><h3>Methods</h3><div>Patients were divided into two datasets: dataset I (January 2020–December 2021) for model development and dataset II (January 2022–December 2023) for validation. A diffusion model was applied to address data imbalance. Eleven machine-learning methods, including Random Forest (RF), Logistic Regression, CatBoost, eXtreme Gradient Boosting Light Gradient Boosting Machine, K-Nearest Neighbors Support Vector Machine, Multi-Layer Perceptron, and Gaussian Naive Bayes, and 2 ensemble learning models, were constructed to predict readmissions. Bayesian optimization was used to fine-tune the hyperparameters of these models. Model performance was primarily evaluated using the area under the receiver operating characteristic curve (AUC). Shapley Additive Explanations (SHAP) were utilized to identify and interpret the significance of predictive variables.</div></div><div><h3>Results</h3><div>Dataset I included 489 patients, while dataset II comprised 418 patients, with readmission rates of 15.3 % and 16.0 %, respectively. The RF model achieved the highest predictive performance (AUC = 0.9116, sensitivity = 0.8806, specificity = 0.7806). SHAP analysis identified readiness for hospital discharge as the most significant predictor of readmission.</div></div><div><h3>Conclusion</h3><div>The RF model shows promise for predicting unplanned 30-day readmissions in older patients with IS. Multi-center studies with larger sample sizes are needed to validate these findings.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"197 ","pages":"Article 105845"},"PeriodicalIF":3.7,"publicationDate":"2025-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143508973","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Explainability and uncertainty: Two sides of the same coin for enhancing the interpretability of deep learning models in healthcare","authors":"Massimo Salvi ,&nbsp;Silvia Seoni ,&nbsp;Andrea Campagner ,&nbsp;Arkadiusz Gertych ,&nbsp;U.Rajendra Acharya ,&nbsp;Filippo Molinari ,&nbsp;Federico Cabitza","doi":"10.1016/j.ijmedinf.2025.105846","DOIUrl":"10.1016/j.ijmedinf.2025.105846","url":null,"abstract":"<div><h3>Background</h3><div>The increasing use of Deep Learning (DL) in healthcare has highlighted the critical need for improved transparency and interpretability. While Explainable Artificial Intelligence (XAI) methods provide insights into model predictions, reliability cannot be guaranteed by simply relying on explanations.</div></div><div><h3>Objectives</h3><div>This position paper proposes the integration of Uncertainty Quantification (UQ) with XAI methods to improve model reliability and trustworthiness in healthcare applications.</div></div><div><h3>Methods</h3><div>We examine state-of-the-art XAI and UQ techniques, discuss implementation challenges, and suggest solutions to combine UQ with XAI methods. We propose a framework for estimating both aleatoric and epistemic uncertainty in the XAI context, providing illustrative examples of their potential application.</div></div><div><h3>Results</h3><div>Our analysis indicates that integrating UQ with XAI could significantly enhance the reliability of DL models in practice. This approach has the potential to reduce interpretation biases and over-reliance, leading to more cautious and conscious use of AI in healthcare.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"197 ","pages":"Article 105846"},"PeriodicalIF":3.7,"publicationDate":"2025-02-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143474963","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Electrocardiographic-Driven artificial intelligence Model: A new approach to predicting One-Year mortality in heart failure with reduced ejection fraction patients","authors":"Hak Seung Lee ,&nbsp;Ga In Han ,&nbsp;Kyung-Hee Kim ,&nbsp;Sora Kang ,&nbsp;Jong-Hwan Jang ,&nbsp;Yong-Yeon Jo ,&nbsp;Jeong Min Son ,&nbsp;Min Sung Lee ,&nbsp;Joon-myoung Kwon ,&nbsp;Byung-Hee Oh","doi":"10.1016/j.ijmedinf.2025.105843","DOIUrl":"10.1016/j.ijmedinf.2025.105843","url":null,"abstract":"<div><h3>Background</h3><div>Despite the proliferation of heart failure (HF) mortality prediction models, their practical utility is limited. Addressing this, we utilized a significant dataset to develop and validate a deep learning artificial intelligence (AI) model for predicting one-year mortality in heart failure with reduced ejection fraction (HFrEF) patients. The study’s focus was to assess the effectiveness of an AI algorithm, trained on an extensive collection of ECG data, in predicting one-year mortality in HFrEF patients.</div></div><div><h3>Methods</h3><div>We selected HFrEF patients who had high-quality baseline ECGs from two hospital visits between September 2016 and May 2021. A total of 3,894 HFrEF patients (64% male, mean age 64.3, mean ejection fraction 29.8%) were included. Using this ECG data, we developed a deep learning model and evaluated its performance using the area under the receiver operating characteristic curve (AUROC).</div></div><div><h3>Results</h3><div>The model, validated against 16,228 independent ECGs from the original cohort, achieved an AUROC of 0.826 (95 % CI, 0.794–0.859). It displayed a high sensitivity of 99.0 %, positive predictive value of 16.6 %, and negative predictive value of 98.4 %. Importantly, the deep learning algorithm emerged as an independent predictor of 1-yr mortality of HFrEF patients with an adjusted hazards ratio of 4.12 (95 % CI 2.32–7.33, p &lt; 0.001).</div></div><div><h3>Conclusions</h3><div>The depth and quality of our dataset and our AI-driven ECG analysis model significantly enhance the prediction of one-year mortality in HFrEF patients. This promises a more personalized, future-focused approach in HF patient management.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"197 ","pages":"Article 105843"},"PeriodicalIF":3.7,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143454558","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing security in patient medical information exchange: A qualitative study","authors":"Aurélia Manns ,&nbsp;Thomas Pezziardi ,&nbsp;Natacha Kadlub ,&nbsp;Anita Burgun ,&nbsp;Alban Destrez ,&nbsp;Rosy Tsopra","doi":"10.1016/j.ijmedinf.2025.105841","DOIUrl":"10.1016/j.ijmedinf.2025.105841","url":null,"abstract":"<div><h3>Background</h3><div>The digital transition has changed the practice of exchanging patient medical information between health professionals. Challenges include the involvement of multiple professionals with varying communication styles, the exponential growth of diverse data types, interoperability issues due to non-integrated tools, and heightened security risks stemming from the use of unsecured applications and personal devices.</div><div>Here, we aimed to understand how to help health surgeons to better consider security during data exchange.</div></div><div><h3>Methods</h3><div>We conducted a qualitative research with 20 interviews with surgeons working in wards of several French institutions. The verbatims were analyzed manually by two researchers using an iterative thematic approach, resulting in a framework to improve practitioners’ security awareness.</div></div><div><h3>Results</h3><div>Our findings emphasize the necessity of a multifaceted strategy, as a single secure application is not sufficient. Effective solutions require combining tailored digital tools with educational initiatives and institutional support. The proposed application must meet specific requirements; and simultaneously, hospitals must provide clear regulations, financial investment, and continuous support to reduce professional constraints.</div></div><div><h3>Conclusion</h3><div>This study underscores the need for a holistic approach, spanning education, institutional backing, and advanced technology, to enhance data security in healthcare. Future studies could extend our framework by considering other healthcare settings and patient perspectives.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"197 ","pages":"Article 105841"},"PeriodicalIF":3.7,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143454559","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparing physician and large language model responses to influenza patient questions in the online health community","authors":"Hong Wu,&nbsp;Mingyu Li,&nbsp;Li Zhang","doi":"10.1016/j.ijmedinf.2025.105836","DOIUrl":"10.1016/j.ijmedinf.2025.105836","url":null,"abstract":"<div><h3>Introduction</h3><div>During influenza season, some patients tend to seek medical advice through online platforms. However, due to time constraints, the informational and emotional support provided by physicians is limited. Large language models (LLMs) can rapidly provide medical knowledge and empathy, but their capacity for providing informational support to patients with influenza and assisting physicians in providing emotional support is unclear. Therefore, this study evaluated the quality of LLM-generated influenza advice and its emotional support performance in comparison with physician advice.</div></div><div><h3>Methods</h3><div>This study utilized 200 influenza question–answer pairs from the online health community. Data collection consisted of two parts: (1) A panel of board-certified physicians evaluated the quality of LLM advice vs physician advice. (2) Physician advice was polished using an LLM, and the LLM-rewritten advice was compared to the original physician advice using the LLM module.</div></div><div><h3>Results</h3><div>For informational support, there was no significant difference between LLM and physician advice in terms of the presence of incorrect information, omission of information, extent of harm or empathy. Nevertheless, compared to physician advice, LLM advice was more likely to cause harm and to be in line with medical consensus. LLM was also able to assist physicians in providing emotional support, since the LLM-rewritten advice was significantly more respectful, friendly and empathetic, when compared with physician advice. Also, the LLM-rewritten advice was logically smooth. In most cases, LLM did not add or omit the original medical information.</div></div><div><h3>Conclusion</h3><div>This study suggests that LLMs can provide informational and emotional support for influenza patients. This may help to alleviate the pressure on physicians and promote physician-patient communication.</div></div>","PeriodicalId":54950,"journal":{"name":"International Journal of Medical Informatics","volume":"197 ","pages":"Article 105836"},"PeriodicalIF":3.7,"publicationDate":"2025-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143454557","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}