JMIR Medical Informatics最新文献

{"title":"Interpretable Artificial Intelligence Analysis of Functional Magnetic Resonance Imaging for Migraine Classification: Quantitative Study.","authors":"Guohao Li, Hao Yang, Li He, Guojun Zeng","doi":"10.2196/72155","DOIUrl":"10.2196/72155","url":null,"abstract":"<p><strong>Background: </strong>Deep learning has demonstrated significant potential in advancing computer-aided diagnosis for neuropsychiatric disorders, such as migraine, enabling patient-specific diagnosis at an individual level. However, despite the superior accuracy of deep learning models, the interpretability of image classification models remains limited. Their black-box nature continues to pose a major obstacle in clinical applications, hindering biomarker discovery and personalized treatment.</p><p><strong>Objective: </strong>This study aims to investigate explainable artificial intelligence (XAI) techniques combined with multiple functional magnetic resonance imaging (fMRI) indicators to (1) compare their efficacy in migraine classification, (2) identify optimal model-indicator pairings, and (3) evaluate XAI's potential in clinical diagnostics by localizing discriminative brain regions.</p><p><strong>Methods: </strong>We analyzed resting-state fMRI data from 64 participants, including 21 (33%) patients with migraine without aura, 15 (23%) patients with migraine with aura, and 28 (44%) healthy controls. Three fMRI metrics-amplitude of low-frequency fluctuation, regional homogeneity, and regional functional connectivity strength (RFCS)-were extracted and classified using GoogleNet, ResNet18, and Vision Transformer. For comprehensive model comparison, conventional machine learning methods, including support vector machine and random forest, were also used as benchmarks. Model performance was evaluated through accuracy and area under the curve metrics, while activation heat maps were generated via gradient-weighted class activation mapping for convolutional neural networks and self-attention mechanisms for Vision Transformer.</p><p><strong>Results: </strong>The GoogleNet model combined with RFCS indicators achieved the best classification performance, with an accuracy of >98.44% and an area under the receiver operating characteristic curve of 0.99 for the test set. In addition, among the 3 indicators, the RFCS indicator improved accuracy by approximately 8% compared with the amplitude of low-frequency fluctuation. Brain activation heat maps generated by XAI technology revealed that the precuneus and cuneus were the most discriminative brain regions, with slight activation also observed in the frontal gyrus.</p><p><strong>Conclusions: </strong>The use of XAI technology combined with brain region features provides visual explanations for the progression of migraine in patients. Understanding the decision-making process of the network has significant potential for clinical diagnosis of migraines, offering promising applications in enhancing diagnostic accuracy and aiding in the development of new diagnostic techniques.</p>","PeriodicalId":56334,"journal":{"name":"JMIR Medical Informatics","volume":"13 ","pages":"e72155"},"PeriodicalIF":3.8,"publicationDate":"2025-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12444220/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144994087","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":"Resilience in the Face of Disruption: Viewpoint on the CrowdStrike Incident in July 2024.","authors":"Christopher R Dennis, Christopher S Evans, Kathleen Duckworth, Misty McLawhorn Skinner, John Hanna, Tanya Thompson, Donette Herring, Richard J Medford","doi":"10.2196/69958","DOIUrl":"10.2196/69958","url":null,"abstract":"<p><strong>Unlabelled: </strong>In an era where health care is increasingly dependent on digital infrastructure, the resilience of health IT systems has become a cornerstone of patient safety and operational continuity. As cyber threats grow in frequency and sophistication, health care organizations have turned to advanced cybersecurity tools to safeguard their systems. Yet even the most robust defenses can falter. On July 19, 2024, a routine update from a widely used cybersecurity platform triggered a widespread IT disruption. A flawed sensor configuration led to 8647 \"blue screen of death\" (BSOD) events, with 729 devices requiring manual remediation. What unfolded was not just a technical crisis but a test of organizational agility, collaboration, and resilience. This viewpoint traces the response to that disruption, highlighting the pivotal role of clinical informaticists and the coordinated efforts that enabled a rapid recovery. From the formation of an incident response team to the triage and mitigation of impacted systems, the response was swift and strategic. Clinical informaticists emerged as key players, bridging the gap between technical teams and frontline care providers. They identified workflow disruptions, facilitated communication, and ensured that patient care remained as uninterrupted as possible. Despite the scale of the outage, operations continued with minimal disruption-thanks to early recognition, decisive action, and cross-disciplinary collaboration. This incident underscored the importance of a well-practiced response plan, clear communication channels, and the integration of clinical expertise in technical recovery efforts. As we reflect on this event, several lessons emerge: the need for continuous refinement of incident response strategies, the value of regular training exercises, and the critical role of clinical informatics in navigating digital crises. This paper calls for a renewed commitment to building resilient health IT ecosystems-ones that can withstand disruption and continue to support the delivery of safe, effective care.</p>","PeriodicalId":56334,"journal":{"name":"JMIR Medical Informatics","volume":"13 ","pages":"e69958"},"PeriodicalIF":3.8,"publicationDate":"2025-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12404578/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980461","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":"Deep Learning Radiomics Model Based on Computed Tomography Image for Predicting the Classification of Osteoporotic Vertebral Fractures: Algorithm Development and Validation.","authors":"Jiayi Liu, Lincen Zhang, Yousheng Yuan, Jun Tang, Yongkang Liu, Liang Xia, Jun Zhang","doi":"10.2196/75665","DOIUrl":"https://doi.org/10.2196/75665","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Osteoporotic vertebral fractures (OVFs) are common in older adults and often lead to disability if not properly diagnosed and classified. With the increased use of computed tomography (CT) imaging and the development of radiomics and deep learning technologies, there is potential to improve the classification accuracy of OVFs.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Objective: &lt;/strong&gt;This study aims to evaluate the efficacy of a deep learning radiomics model, derived from CT imaging, in accurately classifying OVFs.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;The study analyzed 981 patients (aged 50-95 years; 687 women, 294 men), involving 1098 vertebrae, from 3 medical centers who underwent both CT and magnetic resonance imaging examinations. The Assessment System of Thoracolumbar Osteoporotic Fractures (ASTLOF) classified OVFs into Classes 0, 1, and 2. The data were categorized into 4 cohorts: training (n=750), internal validation (n=187), external validation (n=110), and prospective validation (n=51). Deep transfer learning used the ResNet-50 architecture, pretrained on RadImageNet and ImageNet, to extract imaging features. Deep transfer learning-based features were combined with radiomics features and refined using Least Absolute Shrinkage and Selection Operator (LASSO) regression. The performance of 8 machine learning classifiers for OVF classification was assessed using receiver operating characteristic metrics and the \"One-vs-Rest\" approach. Performance comparisons between RadImageNet- and ImageNet-based models were performed using the DeLong test. Shapley Additive Explanations (SHAP) analysis was used to interpret feature importance and the predictive rationale of the optimal fusion model.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;Feature selection and fusion yielded 33 and 54 fused features for the RadImageNet- and ImageNet-based models, respectively, following pretraining on the training set. The best-performing machine learning algorithms for these 2 deep learning radiomics models were the multilayer perceptron and Light Gradient Boosting Machine (LightGBM). The macro-average area under the curve (AUC) values for the fused models based on RadImageNet and ImageNet were 0.934 and 0.996, respectively, with DeLong test showing no statistically significant difference (P=2.34). The RadImageNet-based model significantly surpassed the ImageNet-based model across internal, external, and prospective validation sets, with macro-average AUCs of 0.837 versus 0.648, 0.773 versus 0.633, and 0.852 versus 0.648, respectively (P&lt;.05). Using the binary \"One-vs-Rest\" approach, the RadImageNet-based fused model achieved superior predictive performance for Class 2 (AUC=0.907, 95% CI 0.805-0.999), with Classes 0 and 1 following (AUC/accuracy=0.829/0.803 and 0.794/0.768, respectively). SHAP analysis provided a visualization of feature importance in the RadImageNet-based fused model, highlighting the top 3 most influential features: cluster shade, mean, and large area low gr","PeriodicalId":56334,"journal":{"name":"JMIR Medical Informatics","volume":"13 ","pages":"e75665"},"PeriodicalIF":3.8,"publicationDate":"2025-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12396830/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980161","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 Machine Learning Approach for Identifying People With Neuroinfectious Diseases in Electronic Health Records: Algorithm Development and Validation.","authors":"Arjun Singh, Shadi Sartipi, Haoqi Sun, Rebecca Milde, Niels Turley, Carson Quinn, G Kyle Harrold, Rebecca L Gillani, Sarah E Turbett, Sudeshna Das, Sahar Zafar, Marta Fernandes, M Brandon Westover, Shibani S Mukerji","doi":"10.2196/63157","DOIUrl":"https://doi.org/10.2196/63157","url":null,"abstract":"<p><strong>Background: </strong>Identifying neuroinfectious disease (NID) cases using International Classification of Diseases billing codes is often imprecise, while manual chart reviews are labor-intensive. Machine learning models can leverage unstructured electronic health records to detect subtle NID indicators, process large data volumes efficiently, and reduce misclassification. While accurate NID classification is needed for research and clinical decision support, using unstructured notes for this purpose remains underexplored.</p><p><strong>Objective: </strong>The objective of this study is to develop and validate a machine learning model to identify NIDs from unstructured patient notes.</p><p><strong>Methods: </strong>Clinical notes from patients who had undergone lumbar puncture were obtained using the electronic health record of an academic hospital network (Mass General Brigham [MGB]), with half associated with NID-related diagnostic codes. Ground truth was established by chart review with 6 NID-expert physicians. NID keywords were generated with regular expressions, and extracted texts were converted into bag-of-words representations using n-grams (n=1, 2, 3). Notes were randomly split into training (80%), 2400 notes out of 3000, and hold-out testing (20%), 600 notes out of 3000, sets. Feature selection was performed using logistic regression with L1 regularization. An extreme gradient boosting (XGBoost) model classified NID cases, and performance was evaluated using the area under the receiver operating curve (AUROC) and the precision-recall curve (AUPRC). The performance of the natural language processing (NLP) model was contrasted with the Llama 3.2 auto-regressive model on the MGB test set. The NLP model was additionally validated on external data from an independent hospital (Beth Israel Deaconess Medical Center [BIDMC]).</p><p><strong>Results: </strong>This study included 3000 patient notes from MGB from January 22, 2010, to September 21, 2023. Of 1284 initial n-gram features, 342 were selected, with the most significant features being \"meningitis,\" \"ventriculitis,\" and \"meningoencephalitis.\" The XGBoost model achieved an AUROC of 0.98 (95% CI 0.96-0.99) and AUPRC of 0.89 (95% CI 0.83-0.94) on MGB test data. In comparison, NID identification using International Classification of Diseases billing codes showed high sensitivity (0.97) but poor specificity (0.59), overestimating NID cases. Llama 3.2 improved specificity (0.94) but had low sensitivity (0.64) and an AUROC of 0.80. In contrast, our NLP model balanced specificity (0.96) and sensitivity (0.84), outperforming both methods in accuracy and reliability on MGB data. When tested on external data from BIDMC, the NLP model maintained an AUROC of 0.98 (95% CI 0.96-0.99), with an AUPRC of 0.78 (95% CI 0.66-0.89).</p><p><strong>Conclusions: </strong>The NLP model accurately identifies NID cases from clinical notes. Validated across 2 independent hospital datasets, the model demons","PeriodicalId":56334,"journal":{"name":"JMIR Medical Informatics","volume":"13 ","pages":"e63157"},"PeriodicalIF":3.8,"publicationDate":"2025-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12396800/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980435","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":"Patient-Centered Outpatient Process Optimization System Based on Intelligent Guidance in a Large Tertiary Hospital in China: Implementation Report.","authors":"Xiaoyi Wang, Rujia Zhang, Zhihong Gao, Mengxuan Xia, Songjia Zhang, Lizheng Ge, Yuyang Zhu, Haojie Jin, Shenglian Pan, Manman Zheng, Chun Chen, Xiangyang Zhang","doi":"10.2196/60219","DOIUrl":"https://doi.org/10.2196/60219","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;In large tertiary hospitals across China, outpatient patients often encounter the \"three longs and one short\" (long registration time, long waiting time, long time for medicine collection, and short time for medical treatment) phenomenon. This scenario contributes to suboptimal patient experiences and declining satisfaction with health care services. To address the issue of long waiting times, many hospitals in China have implemented a range of measures. However, these measures have only improved individual aspects of the patient experience, with limited overall impact. Currently, there is a lack of comprehensive, intelligent reform for the entire patient service process in the medical system. Therefore, there is an urgent need to integrate and optimize the entire patient service process, providing real-time intelligent guidance within hospitals. This would help reduce waiting times for patients and enhance their satisfaction.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Objective: &lt;/strong&gt;This study aims to introduce a patient-centered intelligent guidance system and report on the impact of its implementation on outpatient waiting times and patient satisfaction in hospitals.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;The intelligent guidance system was designed with a patient-centered approach, leveraging internet and big data technologies. The system seamlessly connects various steps of the outpatient medical process, facilitating functions including automated check-in and comprehensive intelligent guidance for patients' medical visits, thus enhancing the efficiency and quality of health care delivery. This system has been implemented in a tertiary hospital in China. To assess the system's effectiveness, we compared outpatient visit data, waiting time data, and patient satisfaction levels between the preimplementation and postimplementation periods from 2019 to 2022. We analyzed the changes in patients' average waiting times and satisfaction levels after the system was implemented.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;One year after the introduction of the intelligent guidance system, the number of outpatient visits increased from 5,067,958 to 5,456,151. The waiting time for outpatient patients was significantly reduced. The waiting time for consultation decreased by 2.84 minutes (mean 41.14, SD 2.31 min vs mean 38.30, SD 1.89 min; P&lt;.001). The waiting time for examination decreased by 3.35 minutes (mean 47.83, SD 1.10 min vs mean 44.48, SD 1.67 min; P&lt;.001). Consultation time increased to 3.43 minutes (mean 2.85, SD 0.03 min vs mean 3.43, SD 0.26 min; P&lt;.001). After the system was launched, patient satisfaction increased from 89.99% (SD 2.78%) in 2021 to 92.72% (SD 0.18%) in 2022 (P=.005).&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions: &lt;/strong&gt;The patient-centered intelligent guidance system reported in this study proved beneficial for tertiary medical institutions striving to alleviate the outpatient burden caused by prolonged waiting times. Through continuous transformation","PeriodicalId":56334,"journal":{"name":"JMIR Medical Informatics","volume":"13 ","pages":"e60219"},"PeriodicalIF":3.8,"publicationDate":"2025-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12396732/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980430","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":"Natural Language Processing and <i>ICD-10</i> Coding for Detecting Bleeding Events in Discharge Summaries: Comparative Cross-Sectional Study.","authors":"Frederic Gaspar, Mehdi Zayene, Claire Coumau, Elliott Bertrand, Marie Bettex, Marie Annick Le Pogam, Chantal Csajka","doi":"10.2196/67837","DOIUrl":"https://doi.org/10.2196/67837","url":null,"abstract":"&lt;p&gt;&lt;strong&gt;Background: &lt;/strong&gt;Bleeding adverse drug events (ADEs), particularly among older inpatients receiving antithrombotic therapy, represent a major safety concern in hospitals. These events are often underdetected by conventional rule-based systems relying on structured electronic medical record data, such as the ICD-10 (International Statistical Classification of Diseases and Related Health Problems 10th Revision) codes, which lack the granularity to capture nuanced clinical narratives.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Objective: &lt;/strong&gt;This study aimed to develop and evaluate a natural language processing (NLP) model to detect and categorize bleeding ADEs in discharge summaries of older adults. Specifically, the model was designed to distinguish between \"clinically significant bleeding,\" \"severe bleeding,\" \"history of bleeding,\" and \"no bleeding,\" and was compared with a rule-based algorithm using ICD-10 codes.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Methods: &lt;/strong&gt;Clinicians manually annotated 400 discharge summaries, comprising 65,706 sentences, into four categories: \"no bleeding,\" \"clinically significant bleeding,\" \"severe bleeding,\" and \"history of bleeding.\" The dataset was divided into a training set (70%, 47,100 sentences) and a test set (30%, 18,606 sentences). Two detection approaches were developed and evaluated: (1) an NLP model using binary logistic regression and support vector machine classifiers, and (2) a traditional rule-based algorithm relying exclusively on predefined ICD-10 codes. To address class imbalance, with most sentences categorized as irrelevant (\"no bleeding\"), a class-weighting strategy was applied in the NLP model. Model performance was assessed using accuracy, precision, recall, F1-score, and receiver operating characteristic (ROC) curve analyses, with manual annotations as the gold standard.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Results: &lt;/strong&gt;The NLP model significantly outperformed the rule-based approach across all evaluation metrics. At the document level, the NLP model achieved macro-average scores of 0.81 for accuracy and 0.80 for F1-score. Precision was particularly high for detecting severe (0.92) and clinically significant bleeding events (0.87), demonstrating strong classification capability despite class imbalance. ROC analyses confirmed the model's robust diagnostic performance, yielding an area under the curve (AUC) of 0.91 when distinguishing irrelevant sentences from potential bleeding events, 0.88 for identifying historical mentions of bleeding, and notably, 0.94 for differentiating clinically significant from severe bleeding. In contrast, the rule-based ICD-10 model demonstrated high precision (0.94) for clinically significant bleeding but poor recall (0.03) for severe bleeding events, reflecting frequent missed detections. This limitation arose due to its reliance on commonly used ICD-10 codes (eg, gastrointestinal hemorrhage) and inadequate capture of rare severe bleeding conditions such as shock due to hemorrhage.&lt;/p&gt;&lt;p&gt;&lt;strong&gt;Conclusions: &lt;/","PeriodicalId":56334,"journal":{"name":"JMIR Medical Informatics","volume":"13 ","pages":"e67837"},"PeriodicalIF":3.8,"publicationDate":"2025-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12396801/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980396","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 Machine Learning Algorithm With an Oversampling Technique in Limited Data Scenarios for the Prediction of Present and Future Restorative Treatment Need: Development and Validation Study.","authors":"Elina Väyrynen, Otso Tirkkonen, Henna Tiensuu, Jaakko Suutala, Vuokko Anttonen, Marja-Liisa Laitala, Katri Kukkola, Saujanya Karki","doi":"10.2196/75117","DOIUrl":"10.2196/75117","url":null,"abstract":"<p><strong>Background: </strong>Untreated dental caries is the most common health condition worldwide. Therefore, new strategies need to be developed to reduce the manifestations of dental caries.</p><p><strong>Objective: </strong>This study aimed to develop and test a machine learning (ML) algorithm for detecting present and predicting future carious lesions in the adolescent population using a set of easy-to-collect predictive variables. In addition, this study aimed to deal with an imbalanced and small dataset using an oversampling method.</p><p><strong>Methods: </strong>This population-based study was conducted among secondary schoolchildren, aged between 13 and 17 years, from the northern parts of Finland in 2022. After meeting the inclusion criteria, a total of 218 participants were included in this study. The inclusion criteria consisted of participants having completed a web-based risk assessment questionnaire and having undergone a clinical examination at public health care services. Dental caries (International Caries Detection and Assessment System [ICDAS] scores of 4, 5, and 6; ie, ICDAS 4-6) and active initial caries (ICDAS 2+, 3+) were considered as outcomes. Several predictors, such as behavioral and dietary habits, were included. An extreme gradient boosting model was developed, tested, and assessed for its predictive performance. A 4-fold cross-validation was performed using the nested resampling technique. The random oversampling examples method and the k-nearest neighbors classifiers were used for all 4 folds. The mean (SD) performance of all the folds was computed.</p><p><strong>Results: </strong>Dental caries (ICDAS 2+,3+,4-6) were prevalent in 65.6% (143/218) of the participants. The mean area under the curve was 0.77 (SD 0.04) and the mean F₁-score was 0.82 (SD 0.06) for the extreme gradient boosting model. Similarly, the mean area under the curve and mean F₁-scores after oversampling were 0.74 (SD 0.05) and 0.79 (SD 0.04), respectively. The Shapley additive explanation values were calculated for all 4 folds to assess feature importance, revealing that previous dental fillings were the feature most strongly associated with the need for restorative treatment.</p><p><strong>Conclusions: </strong>On the basis of the performance metrics, the ML algorithm developed and tested in this study can be considered good. The ML algorithm could serve as a cost-effective screening tool for dental professionals to identify the risk of future restorative treatment needs. However, future studies with longitudinal cohorts and longitudinal data, along with external validation for generalizability, are needed to validate our model.</p>","PeriodicalId":56334,"journal":{"name":"JMIR Medical Informatics","volume":" ","pages":"e75117"},"PeriodicalIF":3.8,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12426571/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144801057","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":"Development of a Large-Scale Dataset of Chest Computed Tomography Reports in Japanese and a High-Performance Finding Classification Model: Dataset Development and Validation Study.","authors":"Yosuke Yamagishi, Yuta Nakamura, Tomohiro Kikuchi, Yuki Sonoda, Hiroshi Hirakawa, Shintaro Kano, Satoshi Nakamura, Shouhei Hanaoka, Takeharu Yoshikawa, Osamu Abe","doi":"10.2196/71137","DOIUrl":"10.2196/71137","url":null,"abstract":"<p><strong>Background: </strong>Recent advances in large language models have highlighted the need for high-quality multilingual medical datasets. Although Japan is a global leader in computed tomography (CT) scanner deployment and use, the absence of large-scale Japanese radiology datasets has hindered the development of specialized language models for medical imaging analysis. Despite the emergence of multilingual models and language-specific adaptations, the development of Japanese-specific medical language models has been constrained by a lack of comprehensive datasets, particularly in radiology.</p><p><strong>Objective: </strong>This study aims to address this critical gap in Japanese medical natural language processing resources, for which a comprehensive Japanese CT report dataset was developed through machine translation, to establish a specialized language model for structured classification. In addition, a rigorously validated evaluation dataset was created through expert radiologist refinement to ensure a reliable assessment of model performance.</p><p><strong>Methods: </strong>We translated the CT-RATE dataset (24,283 CT reports from 21,304 patients) into Japanese using GPT-4o mini. The training dataset consisted of 22,778 machine-translated reports, and the validation dataset included 150 reports carefully revised by radiologists. We developed CT-BERT-JPN, a specialized Bidirectional Encoder Representations from Transformers (BERT) model for Japanese radiology text, based on the \"tohoku-nlp/bert-base-japanese-v3\" architecture, to extract 18 structured findings from reports. Translation quality was assessed with Bilingual Evaluation Understudy (BLEU) and Recall-Oriented Understudy for Gisting Evaluation (ROUGE) scores and further evaluated by radiologists in a dedicated human-in-the-loop experiment. In that experiment, each of a randomly selected subset of reports was independently reviewed by 2 radiologists-1 senior (postgraduate year [PGY] 6-11) and 1 junior (PGY 4-5)-using a 5-point Likert scale to rate: (1) grammatical correctness, (2) medical terminology accuracy, and (3) overall readability. Inter-rater reliability was measured via quadratic weighted kappa (QWK). Model performance was benchmarked against GPT-4o using accuracy, precision, recall, F1-score, ROC (receiver operating characteristic)-AUC (area under the curve), and average precision.</p><p><strong>Results: </strong>General text structure was preserved (BLEU: 0.731 findings, 0.690 impression; ROUGE: 0.770-0.876 findings, 0.748-0.857 impression), though expert review identified 3 categories of necessary refinements-contextual adjustment of technical terms, completion of incomplete translations, and localization of Japanese medical terminology. The radiologist-revised translations scored significantly higher than raw machine translations across all dimensions, and all improvements were statistically significant (P<.001). CT-BERT-JPN outperformed GPT-4o on 11 of 18 findin","PeriodicalId":56334,"journal":{"name":"JMIR Medical Informatics","volume":"13 ","pages":"e71137"},"PeriodicalIF":3.8,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12392688/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980171","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 Machine Learning Model for Predicting Sarcopenia Among Middle-Aged Adults: Development and External Validation.","authors":"Hye Jin Chong","doi":"10.2196/75760","DOIUrl":"10.2196/75760","url":null,"abstract":"<p><strong>Background: </strong>Sarcopenia is a common muscle disorder in older adults, and its early identification and management in middle-aged populations are essential for ensuring a healthier later life. Detecting sarcopenia at an earlier stage may reduce the future burden on health care systems and enhance the quality of life in older adults. Machine learning (ML) models can evaluate large datasets, identify essential variables, and find complicated correlations between input variables. However, using ML models to detect sarcopenia remains an unsatisfied need.</p><p><strong>Objective: </strong>This study aimed to develop and externally validate an ML model to predict sarcopenia risk among middle-aged adults using a nationally representative dataset.</p><p><strong>Methods: </strong>We analyzed data from 1926 participants aged 40 to 64 years and enrolled in the 2022 Korea National Health and Nutrition Examination Survey (KNHANES). Sarcopenia was diagnosed and defined based on the 2019 Asian Working Group for Sarcopenia criteria, which incorporate both low muscle mass and reduced muscle strength. Muscle mass was assessed using bioelectrical impedance analysis with cutoffs of <7.0 kg/m² for men and <5.7 kg/m² for women. Muscle strength was measured via handgrip strength using a digital dynamometer with thresholds of <28 kg for men and <18 kg for women. Participants meeting both criteria were classified as those with sarcopenia. Four ML algorithms, random forest, support vector machine, extreme gradient boosting, and logistic regression, were used to identify risk factors of sarcopenia and predict its likelihood. The top-performing model was subsequently validated in an external cohort of 2247 middle-aged adults from the 2023 KNHANES. Model performance was assessed using the F₂-score, area under the curve of a receiver operating characteristic curve, and sensitivity. All analyses were conducted using Python 3.13.2 (Python Software Foundation).</p><p><strong>Results: </strong>Among the 4 models, the logistic regression model demonstrated the strongest performance, yielding an area under the curve of 0.85, a sensitivity of 0.92, and an F₂-score of 0.66. External validation using the 2023 KNHANES dataset confirmed the model's robust performance, indicating its potential for widespread applications.</p><p><strong>Conclusions: </strong>This study developed and externally validated an ML model that accurately identified sarcopenia in middle-aged adults. Leveraging data from a comprehensive national survey, our findings underscore the significance of early detection and customized interventions in midlife to mitigate sarcopenia risk and optimize long-term health outcomes.</p>","PeriodicalId":56334,"journal":{"name":"JMIR Medical Informatics","volume":"13 ","pages":"e75760"},"PeriodicalIF":3.8,"publicationDate":"2025-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12423610/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980358","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":"An Extraction Tool for Venous Thromboembolism Symptom Identification in Primary Care Notes to Facilitate Electronic Clinical Quality Measure Reporting: Algorithm Development and Validation Study.","authors":"John Novoa-Laurentiev, Mica Bowen, Avery Pullman, Wenyu Song, Ania Syrowatka, Jin Chen, Michael Sainlaire, Frank Chang, Krissy Gray, Purushottam Panta, Luwei Liu, Khalid Nawab, Shadi Hijjawi, Richard Schreiber, Li Zhou, Patricia C Dykes","doi":"10.2196/63720","DOIUrl":"https://doi.org/10.2196/63720","url":null,"abstract":"<p><strong>Background: </strong>Diagnosis of venous thromboembolism (VTE) is often delayed, and facilitating earlier diagnosis may improve associated morbidity and mortality. Clinical notes contain information not found elsewhere in the medical record that could facilitate timely VTE diagnosis and accurate quality measurement. However, extracting relevant information from unstructured clinical notes is complex. Today, there are relatively few electronic clinical quality measures (eCQMs) in our national payment program and none that use natural language processing (NLP) techniques for data extraction. NLP holds great promise for making quality measurement more accurate and more efficient. Given the potential of NLP-based applications to facilitate more accurate VTE detection, primary care is one clinical setting in urgent need of this type of tool.</p><p><strong>Objective: </strong>This study aimed to develop a tool that extracts VTE symptoms from clinical notes for use within an eCQM to quantify the rate of delayed diagnosis of VTE in primary care settings.</p><p><strong>Methods: </strong>We iteratively developed an NLP-based data extraction tool, venous thromboembolism symptom extractor (VTExt), on an internal dataset using a rule-based approach to extract VTE symptoms from primary care clinical note text. The VTE symptoms lexicon was derived and optimized with physician guidance and externally validated using datasets from 2 independent health care organizations. We performed 26 rounds of performance evaluation of notes sampled from the case cohort (17,585 patient progress note sentences from 279 patient notes), and 5 rounds of evaluation of the control cohort (2838 patient progress note sentences from 50 patient notes). VTExt's performance was evaluated using evaluation metrics, including area under the curve, positive predictive value, negative predictive value, sensitivity, and specificity.</p><p><strong>Results: </strong>VTExt achieved near-perfect performance in extracting VTE symptoms from primary care notes sampled from records of patients diagnosed with or without VTE. In external validation, VTExt achieved promising performance in 2 additional geographically distant organizations using different electronic health record systems. When compared against a deep learning model and 4 machine learning models, VTExt exhibited similar or even improved performance across all metrics.</p><p><strong>Conclusions: </strong>This study demonstrates a data-driven NLP-based approach to clinical note information extraction that can be generalized to different electronic health record systems across different institutions. Due to the robust performance of this tool, VTExt is the first NLP application to be used in a nationally endorsed eCQM.</p>","PeriodicalId":56334,"journal":{"name":"JMIR Medical Informatics","volume":"13 ","pages":"e63720"},"PeriodicalIF":3.8,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12387394/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980356","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}