Journal of Medical Systems最新文献

{"title":"MEL-IA: An Interoperable AI System for Multimodal Skin Lesion Classification in Hospital Settings.","authors":"Pablo Candela Córcoles, Alberto De Ramón Fernández, Marcelo Saval Calvo, Jose María Salinas Serrano, Diego Guijarro Peral, Daniel Ruiz Fernández","doi":"10.1007/s10916-026-02424-y","DOIUrl":"10.1007/s10916-026-02424-y","url":null,"abstract":"<p><p>Skin cancer represents a growing healthcare challenge, motivating the development of AI sytems that can support lesion assessment while integrating efficiently into clinical workflows. This study presents MEL-IA (MobilE skin Lesion dIAgnosis), an interoperable AI system designed for automated skin lesion classification and full integration with hospital information systems.</p><p><strong>Methods: </strong>The system combines dermatoscopic images and structured clinical metadata using an EfficientNet-B4 - based multimodal model trained on the ISIC 2019 BCN_20000 and MSK subsets. Internal robustness was assessed through stratified 5-fold cross-validation, and external generalization was evaluated on the independent HAM10000 dataset. Interoperability and operational performance were validated through deployment in a real hospital environment using HL7, DICOM, PACS, and HIS/RIS systems.</p><p><strong>Results: </strong>We evaluated internal robustness using stratified five‑fold cross‑validation, with global accuracy of 0.86, macro F1-score of 0.85, and AUC values above 0.97. The external evaluation showed that the model generalizes well under different acquisition conditions, achieving an accuracy of 0.84 and balanced accuracy of 0.65, with sensitivities of 0.60 for melanoma 0.72 for basal cell carcinoma. The multimodal model outperformed the image model baseline across all classes. Technical deployment confirmed full interoperability, > 99% successful study integration, and near real time processing.</p><p><strong>Conclusions: </strong>MEL-IA provides multimodal skin lesion classification capabilities and integrates smoothly into hospital infrastructures based on standards. The results demonstrate technical feasibility, interoperability, and operational viability for deployment within dermatology oriented clinical workflows. Further clinical validation studies are warranted to assess the system's impact on diagnostic decision-making and patient outcomes.</p>","PeriodicalId":16338,"journal":{"name":"Journal of Medical Systems","volume":"50 1","pages":""},"PeriodicalIF":8.8,"publicationDate":"2026-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13260036/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148226124","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Feasibility Evaluation of a Standalone Virtual Reality System for Neonatal Resuscitation Training.","authors":"Shoichiro Amari, Masato Eto, Hiroshi Oyama","doi":"10.1007/s10916-026-02423-z","DOIUrl":"10.1007/s10916-026-02423-z","url":null,"abstract":"<p><p>Maintaining competency in neonatal resuscitation requires repeated practice. However, equipment costs, space, and instructor availability constrain mannequin-based simulation. Standalone virtual reality (VR) training may improve access. This study aimed to evaluate the feasibility, acceptability, and safety of a standalone VR-based neonatal resuscitation training application (NCPR-VR) focused on positive-pressure ventilation, among already-certified perinatal healthcare professionals. This single-arm exploratory feasibility study involved 32 already-certified perinatal healthcare professionals (11 physicians, 14 nurses, and 7 midwives). Feasibility and acceptability were assessed by session completion, self-reported usefulness/recommendation, System Usability Scale (SUS), User Experience Questionnaire (UEQ), and Igroup Presence Questionnaire (IPQ). Safety was monitored by Simulator Sickness Questionnaire (SSQ) and adverse event reporting. Analyses were limited to descriptive statistics. All 32 participants completed the VR session without adverse event-induced discontinuation (completion rate, 100%). Median perceived usefulness and willingness to recommend the application were both 5 (IQR, 4-5) on a 5-point Likert scale. The median SUS score was 65.0 (IQR, 55.0-73.1), indicating moderate usability. UEQ subscales showed favorable ratings for attractiveness, stimulation, and novelty. IPQ involvement was unsatisfactory and experienced realism only satisfactory; presence and spatial presence were acceptable. Simulator sickness was low (median SSQ, 1.9); three participants reported mild symptoms resolving within 10 min. No serious adverse events occurred. NCPR-VR was feasible and acceptable among already-certified providers, with full session completion and no serious adverse events. Involvement and experienced realism require improvement. Trial registration: UMIN000058975 (University hospital Medical Information Network Clinical Trials Registry; registered September 6, 2025).</p>","PeriodicalId":16338,"journal":{"name":"Journal of Medical Systems","volume":"50 1","pages":""},"PeriodicalIF":8.8,"publicationDate":"2026-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13260161/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148218120","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large Language Models vs. Machine Learning on Structured Perioperative Data: Does Model Choice Matter?","authors":"Theodora Wingert, Xuezhi Dong","doi":"10.1007/s10916-026-02422-0","DOIUrl":"10.1007/s10916-026-02422-0","url":null,"abstract":"<p><p>The study by Ko and colleagues provides evidence that large language models may achieve modestly improved performance compared with traditional machine learning models when applied to structured perioperative data. However, whether such incremental model performance gains can be translated into improvements in clinical decision-making, workflow integration, or patient outcomes remains an important question for future investigation.</p>","PeriodicalId":16338,"journal":{"name":"Journal of Medical Systems","volume":"50 1","pages":""},"PeriodicalIF":8.8,"publicationDate":"2026-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148204855","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Vitabel: A Python Framework for Visualizing and Labelling High-Resolution Physiological Data for Critical Care Machine Learning.","authors":"Simon Orlob, Wolfgang J Kern, Benjamin Hackl, Jan Wnent, Jan-Thorsten Gräsner, Martin Holler","doi":"10.1007/s10916-026-02417-x","DOIUrl":"10.1007/s10916-026-02417-x","url":null,"abstract":"<p><p>Artificial intelligence offers great opportunities in critical care, particularly when a vast amount of continuously acquired physiological data is incorporated. High-quality, reliably labelled data are paramount for developing and training artificial intelligence methods. However, routinely recorded data in critical care are often noisy, and the sheer volume of high-resolution data is challenging to manage. Generalizable solutions for these problems are lacking, restricting progress. To address these barriers, we developed Vitabel, an open-source Python framework for post hoc loading, visualizing, aligning, and annotating medical time series. The framework provides sensible defaults and interactive components for efficient use in preconfigured workflows, while remaining flexible and extendable for custom analysis and annotation pipelines. It integrates seamlessly into Jupyter Notebooks, providing an interactive, customizable interface for visual interaction with the data. In this publication, we demonstrate its utility across three use cases. The code and exemplary data are provided as browser-based demos. Vitabel is freely available and published under the MIT license accompanying this publication.</p>","PeriodicalId":16338,"journal":{"name":"Journal of Medical Systems","volume":"50 1","pages":""},"PeriodicalIF":8.8,"publicationDate":"2026-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13249750/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148204929","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"pedQTNet: A Deep Learning Approach to Estimate Corrected QT Intervals from Multi-Lead Conventional ECG Waveforms in Pediatric Patients.","authors":"Victor M Ruiz, Ivor B Asztalos, Luiz E V Silva, Lingyun Shi, V Ramesh Iyer, Dustin Nash, Victoria L Vetter, Fuchiang Rich Tsui","doi":"10.1007/s10916-026-02386-1","DOIUrl":"10.1007/s10916-026-02386-1","url":null,"abstract":"<p><p>Long QT syndrome (LQTS) is a primary risk factor for ventricular arrhythmias and sudden cardiac death in children. Accurate corrected QT intervals (QTc) measurement is imperative but challenging for non-heart-rhythm specialists, especially in children. We developed and evaluated pedQTNet, a deep neural network model for estimating QTc and detecting LQTS in pediatric patients. We analyzed a cohort of 37,992 patients aged 0-18 years with 65,370 ECGs annotated by pediatric electrophysiologists (PEPs) between 2010 and 2020. Using PEP-annotated QTc measurements as ground truth, pedQTNet was trained and calibrated on raw ECG waveforms to optimize QTc estimation and LQTS classification. Performance was compared to GE Healthcare's Marquette 12SL algorithm, and to PEPs in cross-validation, as well as an additional prospective set of 200 ECGs. In 10-fold cross-validation, pedQTNet estimated QTc's with a mean absolute error (MAE) of 18.8 ms (95% CI: 18.4-19.2) and predicted LQTS at 470 ms with 85% sensitivity (83%-87%), 87% specificity (87%-88%), positive likelihood ratio (PLR) of 6.7 (6.5-7.0), and negative likelihood ratio (NLR) of 0.17 (0.15-0.19), outperforming Marquette 12SL. In the prospective set, pedQTNet had higher sensitivity than PEPs (100% [69%-100%] vs. 71% [53%-85%], P < 0.05), and a lower but not statistically significant NLR (0.00 [0.00-0.70] vs. 0.30 [0.18-0.50], P = 0.2). PedQTNet demonstrated high QTc estimation accuracy and reliable LQTS detection, outperforming a commercial tool and on par with expert interpretation. Its strong performance supports its clinical use for scalable, automated pediatric ECG screening and LQTS risk assessment, offering a practical tool for enhancing pediatric cardiac care.</p>","PeriodicalId":16338,"journal":{"name":"Journal of Medical Systems","volume":"50 1","pages":""},"PeriodicalIF":8.8,"publicationDate":"2026-06-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13233968/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148150011","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Hybrid Deep Learning Framework with Supervised Contrastive Learning for Robust Seizure Detection in Long-Term EEG.","authors":"Haotian Li, Weisen Lu, Xiangwen Zhong, Haozhou Cui, Chuanyu Li, Jiaqi Wang, Zhen Liu, Wei Shang, Weidong Zhou","doi":"10.1007/s10916-026-02395-0","DOIUrl":"10.1007/s10916-026-02395-0","url":null,"abstract":"<p><p>Accurate and timely automatic detection of epileptic seizures is crucial for reducing the workload for visually inspecting long-term electroencephalogram (EEG) and improving diagnostic efficiency. However, this task remains a critical challenge owing to the complex and non-stationary dynamics of ictal EEG activities and the severe class imbalance in long-term EEG data. Therefore, we develop an end-to-end hybrid deep learning framework named CNN-SwT, aiming to detect seizures by integrating Convolutional Neural Network (CNN) and Swin Transformer (SwT). CNN-SwT leverages CNN to capture the spatiotemporal local features from multi-channel EEGs and SwT to further derive their global dependencies. Furthermore, supervised contrastive learning is adopted in place of the traditional cross-entropy loss to address the extreme data imbalance inherent in seizure detection. Through extensive evaluation over the publicly available CHB-MIT long-term EEG database and our SH-SDU database, CNN-SwT achieved an event-based sensitivity of 100% with an FDR of 0.36/h and an event-based sensitivity of 94.44% with an FDR of 0.78/h, respectively. These outstanding cross-database results underscore the robustness of the CNN-SwT model and its potential to advance clinical seizure detection.</p>","PeriodicalId":16338,"journal":{"name":"Journal of Medical Systems","volume":"50 1","pages":""},"PeriodicalIF":8.8,"publicationDate":"2026-06-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148150047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Training for the Digital Clinical Environment: Preliminary Findings in Implementation of Electronic Health Record Platform in Medical Education.","authors":"Ashley Tin, Anthony Wong, Kristy Y Lin, Jack Yang, Avery Buehler, Dongfang Emily Chen, Sounak Dey, Nicholas Stefanko, Christopher Nelson, Aditya Vaidyam","doi":"10.1007/s10916-026-02411-3","DOIUrl":"10.1007/s10916-026-02411-3","url":null,"abstract":"<p><p>Electronic health records (EHRs) are central to modern clinical workflows, yet undergraduate medical education often provides limited hands-on experience with authentic documentation systems. This gap may reduce preparedness for clerkships and residency, limit confidence with chart navigation, and hinder development of digital literacy and clinical reasoning skills. To develop a learner-centered, open-source, web-based educational EHR platform and evaluate its usability, educational value, and curricular integration across preclinical and clinical learners. A browser-based EHR was developed to simulate real-world documentation workflows, support case-based and problem-based learning, and provide structured feedback. The platform enables instructors to create cases, assign activities, and track learner engagement. Fifty-four preclinical medical students participated in preliminary usability testing and completed post session surveys evaluating usability, learning support, and perceived realism using 5-point Likert scales. Responses were analyzed descriptively. Students reported high usability (mean 4.1 ± 0.6) and strong satisfaction with organization, navigation, and learning support. Learners emphasized workflow realism and integration of documentation with communication tasks. Early EHR exposure helped bridge classroom learning and clinical practice. Feedback informed iterative improvements to interface clarity and workflow design. This platform provides scalable, structured exposure to clinical documentation. Early findings suggest improved digital fluency, confidence, and clinical reasoning. Integration of EHR based simulation into curricula may reduce the learning curve associated with real-world EHR use.</p>","PeriodicalId":16338,"journal":{"name":"Journal of Medical Systems","volume":"50 1","pages":""},"PeriodicalIF":8.8,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13226437/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148138231","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DualKG-DC: A Drug-Centric Dual-Layer Knowledge Graph Framework for Drug Combination Prediction.","authors":"Zhenxiang Gao, Scott W Perkins, Satya Parameswaran, Rong Xu","doi":"10.1007/s10916-026-02416-y","DOIUrl":"10.1007/s10916-026-02416-y","url":null,"abstract":"<p><p>Most existing approaches to drug combination discovery are disease-centered, aiming to identify drug pairs for specific diseases. Complementarily, a drug-centered strategy starts from known drug combinations and explores new therapeutic indications, facilitating translational applications by leveraging combinations with established safety profiles. Here, we introduce DualKG-DC, a drug centered computational framework that provides a complementary perspective by identifying potential disease indications for a given drug combination. The dual layer knowledge graph architecture, which is pretrained on a foundation biomedical knowledge graph and subsequently refined on a task specific drug combination subgraph, may reduce reliance on large, labeled datasets by leveraging existing knowledge on drug targets, biological pathways, and observed phenotypic effects. In systematic benchmarking against three state-of-the-art models, DualKG-DC outperformed all comparison models, achieving an average Hits@10 of 0.48, MRR of 0.30, AUROC of 0.99, and AUPRC of 0.31. Notably, in cold start scenarios, DualKG-DC outperformed baseline methods in predicting indications for unseen drug combinations, achieving superior results with an average Hits@10 score of 0.32, an MRR of 0.18, an AUROC of 0.98, and an AUPRC of 0.23. These results highlight DualKG-DC as an effective platform for systematically discovering therapeutic opportunities of drug combinations. By leveraging a dual-layer architecture, the model enables effective knowledge transfer, enhancing predictive performance and robustness, particularly for previously unseen drug combinations.</p>","PeriodicalId":16338,"journal":{"name":"Journal of Medical Systems","volume":"50 1","pages":""},"PeriodicalIF":8.8,"publicationDate":"2026-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13219182/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148042509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Association of the Pay-for-Performance Program with Emergency Service Utilization and Hospital Admissions in Asthma Patients: A National Matched Cohort Study in Taiwan.","authors":"Chang-Hung Tsai, Yeong-Ruey Chu, Pei-Tseng Kung, Wen-Chen Tsai","doi":"10.1007/s10916-026-02415-z","DOIUrl":"10.1007/s10916-026-02415-z","url":null,"abstract":"<p><p>About 339 million people worldwide have asthma, and its prevalence is increasing. The asthma pay-for-performance (P4P) program was implemented, but the effectiveness of P4P programs is still controversial. The study evaluated the relationship between participation in the pay-for-performance (P4P) program and emergency service utilization and hospital admission utilization among asthma patients. This population-based cohort study included those who were newly diagnosed asthma patients from 2011 to 2016 and were followed until the end of 2018. The data were from the National Health Insurance Database of Taiwan. A propensity score matching method was applied to match asthma patients who joined the P4P program to non-P4P patients, ratio 1:1. The Cox proportional hazards model was applied to analyze the relationship between P4P participation status and emergency service utilization and hospital admission utilization. We also conducted alternative analysis using Poisson regression. After matching, 73,117 people who joined and 73,117 who did not join were included for analyses. It was found that asthma patients who joined P4P had a lower risk of emergency service utilization than those who did not join P4P (HR = 0.57, 95% CI = 0.55-0.59). We also found that asthma patients enrolled in the P4P program had a significantly lower risk of hospital admission utilization (HR = 0.58, 95% CI = 0.54-0.62). The Poisson regression analysis showed a high level of consistency with the results mentioned above. Our data showed that participation in the P4P program was associated with emergency service utilization and hospital admission utilization among asthma patients.</p>","PeriodicalId":16338,"journal":{"name":"Journal of Medical Systems","volume":"50 1","pages":""},"PeriodicalIF":8.8,"publicationDate":"2026-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148005298","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine Learning-Predicted Risk Trajectories for Incident Chronic Kidney Disease and Associations with Post-CKD Outcomes.","authors":"Aaron Boussina, Amy M Sitapati, Soo-Young Yoon, Sripad Karne, Jongwoo Seo, Woo-Jung Kim, Hyeon Seok Hwang","doi":"10.1007/s10916-026-02413-1","DOIUrl":"10.1007/s10916-026-02413-1","url":null,"abstract":"<p><p>Chronic kidney disease (CKD) is a progressive condition where risk accumulates before clinical onset. However, strategies for evaluating CKD risk remain limited, and the trajectories of risk over time are poorly understood. We enrolled 51,156 individuals without CKD in the University of California health system between 2012 and 2024. A debiased XGBoost model was developed to predict CKD risk, and the predicted probabilities from this model at each visit were used to derive risk trajectories through latent class growth mixture modeling. The model achieved a time-dependent area under the curve of 0.95 (95% confidence interval (CI), 0.94-0.96), a C-index of 0.93 (0.92-0.94), and a Brier score of 0.027 (0.026-0.028) for predicting 5-year CKD occurrence. Similar predictive performance was observed for 1 to 4 years. Longitudinal patterns of predicted CKD risk revealed three distinct trajectories: gradual (75.9%), progressive (10.1%), and rapid increase (14.0%). Compared to the gradual trajectory, the progressive trajectory was associated with a higher risk of estimated glomerular filtration rate (eGFR) decline of ≥ 25% (adjusted hazard ratio (HR) 1.31; 95% CI, 1.06-1.61) and ≥ 50% (HR 1.59; 95% CI, 1.09-2.31) following CKD onset. The progressive and rapid trajectory was linked to an increased risk of cardiovascular events (HR 1.84; 95% CI 1.18-2.91) and all-cause mortality (HR 2.36; 95% CI 1.49-3.73). Our machine learning model demonstrated strong predictive performance for incident CKD risk and identified three distinct risk trajectories that were associated with different post-CKD outcomes, highlighting the importance of early risk monitoring.</p>","PeriodicalId":16338,"journal":{"name":"Journal of Medical Systems","volume":"50 1","pages":""},"PeriodicalIF":8.8,"publicationDate":"2026-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148005356","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}