JMIR Publications

Recent Articles:

• 

  Machine Learning

  Sep 17, 2025

  An Artificial Intelligence–Based Framework for Predicting Emergency Department Overcrowding: Development and Evaluation Study

  View Authors View Abstract

  Authors List:

  • Orhun Vural  
  • Bunyamin Ozaydin  
  • Khalid Y Aram  
  • James Booth  
  • Brittany Freeman Lindsey  
  • Abdulaziz Ahmed  

  Abstract:

  Background: Emergency department (ED) overcrowding remains a critical challenge, leading to delays in patient care and increased operational strain. Current hospital management strategies often rely on reactive decision-making, addressing congestion only after it occurs. However, effective patient flow management requires early identification of overcrowding risks to allow timely interventions. Machine learning (ML)–based predictive modeling offers a solution by forecasting key patient flow measures, such as waiting count, enabling proactive resource allocation and improved hospital efficiency. Objective: The aim of this study is to develop ML models that predict ED waiting room occupancy (waiting count) at 2 temporal resolutions. The first approach is the hourly prediction model, which estimates the waiting count exactly 6 hours ahead at each prediction time (eg, a 1 PM prediction forecasts 7 PM). The second approach is the daily prediction model, which forecasts the average waiting count for the next 24-hour period (eg, a 5 PM prediction estimates the following day’s average). These predictive tools support resource allocation and help mitigate overcrowding by enabling proactive interventions before congestion occurs. Methods: Data from a partner hospital’s ED in the southeastern United States were used, integrating internal and external sources. Eleven different ML algorithms, ranging from traditional approaches to deep learning architectures, were systematically trained and evaluated on both hourly and daily predictions to determine the models that achieved the lowest prediction error. Experiments optimized feature combinations, and the best models were tested under high patient volume and across different hours to assess temporal accuracy. Results: The best hourly prediction performance was achieved by time series vision transformer plus (TSiTPlus) with a mean absolute error (MAE) of 4.19 and a mean squared error (MSE) of 29.36. The overall hourly waiting count had a mean of 18.11 and a SD (σ) of 9.77. Prediction accuracy varied by time of day, with the lowest MAE at 11 PM (2.45) and the highest at 8 PM (5.45). Extreme case analysis at (mean + 1σ), (mean + 2σ), and (mean + 3σ) resulted in MAEs of 6.16, 10.16, and 15.59, respectively. For daily predictions, an explainable convolutional neural network plus (XCMPlus) achieved the best results with an MAE of 2.00 and a MSE of 6.64. The daily waiting count had a mean of 18.11 and a SD of 4.51. Both models outperformed traditional forecasting approaches across multiple evaluation metrics. Conclusions: The proposed prediction models effectively forecast ED waiting count at both hourly and daily intervals. The results demonstrate the value of integrating diverse data sources and applying advanced modeling techniques to support proactive resource allocation decisions. The implementation of these forecasting tools within hospital management systems has the potential to improve patient flow and reduce overcrowding in emergency care settings. The code is available in our GitHub repository. Trial Registration:

  • JMIR Med Inform 2025;13(1):e73960

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.

• 

  Machine Learning

  Sep 16, 2025

  Using Wearable Device and Machine Learning to Predict Mood Symptoms in Bipolar Disorder: Development and Usability Study

  View Authors View Abstract

  Authors List:

  • Chia-Tung Wu  
  • Ming H Hsieh  
  • I-Ming Chen  
  • Lian-Yin Jhao  
  • Ding-Shan Liu  
  • Ssu-Ming Wang  
  • Chia-Ting Wu  
  • Yi-Ling Chien  

  Abstract:

  Background: Bipolar disorder (BD) is a highly recurrent disorder. Early detection, early intervention, and prevention of recurrent bipolar mood symptoms are key for better prognosis. Objective: This study aims to build prediction models for BD with machine learning algorithms. Methods: This study recruited 24 participants with BD. The Beck Depression Inventory (BDI) and Young Mania Rating Scale (YMRS) were used to evaluate depressive and manic episodes respectively. Using digital biomarkers collected from wearable devices as input, six machine learning algorithms (Logistic Regression, Decision Tree, K-Nearest Neighbors, Random Forest, Adaptive Boosting, and Extreme Gradient Boosting) were used to build predictive models. Results: The prediction model for depressive symptoms achieved 83% accuracy, 0.89 Area Under the Receiver Operating Characteristic curve (AUROC), and 0.65 F1 score on testing data. The prediction model for manic symptoms achieved 91% accuracy, 0.88 AUROC, and 0.25 F1 score on testing data. With the interpretable model Shapely Additive exPlanations (SHAP), we found that relatively high resting heart rate, low activity, and lack of sleep may predict depressive symptoms. Conclusions: This study demonstrated that digital biomarkers could be used to predict depressive and manic symptoms. This prediction model may be beneficial for the early detection of mood symptoms, facilitating timely treatment and helping to prevent BD recurrence.

  • JMIR Med Inform 2025;13(1):e66277

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.

• 

  Advanced Data Analytics in eHealth

  Sep 16, 2025

  Holistic Influence of Multimodal Medical Crowdfunding Affordances on Charitable Crowdfunding Outcome: Systematic Multimodel Analysis Study

  View Authors View Abstract

  Authors List:

  • Yuxuan Du  
  • Yujin Yang  
  • Zihe Li  
  • Jiaolong Xue  

  Abstract:

  Background: Medical crowdfunding has emerged as a critical tool to alleviate the financial burden of health care costs, particularly in regions where economic disparities limit access to medical treatment. Despite its potential, the success rates of medical crowdfunding projects remain low, with only 9% achieving their fundraising goals in China. Previous research has examined isolated factors influencing success, but a holistic understanding of how multimodal affordances—narrativity, visibility, and progress—collectively impact donor behavior and project outcomes is lacking. Objective: This study aims to investigate how medical crowdfunding affordances, as an integrated system, influence the success of charitable crowdfunding projects. Specifically, it explores the roles of narrativity (textual elements), visibility (visual elements), and progress (dynamic updates) affordances, and how these interact with patient demographics to shape donor engagement and fundraising outcomes. Methods: A multimodal analysis was conducted using 1261 medical crowdfunding projects from the Shuidichou platform in China. Machine learning techniques (eg, sentiment analysis via SnowNLP) and regression models were used to examine textual content, visual elements, and progress updates. Control variables included patient age, gender, and beneficiary type. Hypotheses were tested using both continuous (success ratio) and binary (success indicator) measures of project success. In total, 6 models were constructed to examine the influences of affordances. Results: The study found that narrativity affordances—longer titles (model 1a: P=.04; model 3a: P=.03) and detailed surplus fund descriptions (P=.03)—boosted success, while overly lengthy surplus fund explanations had diminishing returns (P=.005). Disease mentions in titles increased donations (model 1a: P=.01; model 3a: P=.003). A neutral tone in the project plan also improved success (P<.001). For visibility affordances, a moderate number of progress photos maximized project success, while excessive visuals reduced impact (P<.001). Progress affordances followed a similar pattern, with a moderate number of updates enhancing success (P<.001). Critically, when all affordances were considered, only progress update frequency retained a strong inverted U-shaped effect on success (P<.001). Demographics, particularly age, also influenced donations: patients at both ends of the age spectrum received greater support , while middle-aged individuals received less (model 1b: P=.02; model 2b: P=.005; model 3b: P=.02). Conclusions: This study advances medical crowdfunding affordance theory by demonstrating the interconnected effects of narrativity, visibility, and progress affordances on project success. Practically, results highlight the importance of strategically crafted titles, targeted demographic disclosures, and balanced progress updates—with moderate update frequency being crucial when controlling all affordances—to enhance donor engagement. Platform designers and project organizers can apply these insights to optimize fundraising outcomes and effectively address health care inequalities. Future research should further investigate visual content analysis and donor psychology to refine engagement strategies. Trial Registration:

  • JMIR Med Inform 2025;13(1):e75563

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.

• 

  Reviews in Medical Informatics

  Sep 16, 2025

  Evidence for the Use of Patient-Reported Outcome Measures in the Treatment of Patients With Noncommunicable Diseases: Systematic Review

  View Authors View Abstract

  Authors List:

  • Marie Villumsen  
  • Benedikte Irene von Osmanski  
  • Kirsten Elisabeth Lomborg  
  • Kirstine Skov Benthien  

  Abstract:

  Background: The use of patient-reported outcome measures (PROMs) as a clinical tool for screening and decision-making has gained widespread interest, with numerous implementation activities across specialties, even though the evidence has not been clear until now. Objective: The aim of this study was to assess the evidence for using PROMs in clinical practice for patients with diabetes, chronic obstructive pulmonary disease (COPD), heart disease, rheumatoid arthritis (RA), and inflammatory bowel disease (IBD). Additionally, we sought to determine the characteristics of the most effective PROM interventions. Methods: We conducted a systematic review of published randomized controlled trials (RCTs) on the use of PROMs for clinical purposes, such as systematic PROM assessment alone or with a predefined PROM-based decision-making method. Eligible studies included adult patients (>18 years) with diabetes, COPD, heart disease, RA, or IBD. We excluded studies using PROMs as an outcome measure or otherwise not meeting the inclusion criteria. We searched the PubMed/MEDLINE, CINAHL, EMBASE, and Web of Science databases until February 2023. Two investigators independently screened titles, abstracts, and relevant full texts. Three investigators completed data extraction and risk-of-bias assessment using version 2 of the Cochrane risk-of-bias tool for randomized trials (RoB 2). The data were presented in a narrative synthesis and in summarized form. Results: The search yielded 21,203 papers, 686 (3.2%) full-text papers were screened, and 56 (8.2%) original studies were included in the review. The studies included patients with heart disease (n=17, 30.4%), COPD (n=13, 23.2%), diabetes (n=10, 17.9%), IBD (n=9, 16.1%), and RA (n=6, 10.7%), as well as patients with mixed diagnoses (n=1, 1.8%). All interventions incorporated systematic PROM assessments. Some interventions additionally used a predefined method for PROM-based decision-making (n=19, 33.9%) or PROM-based dialogue (n=9, 16.1%), while 5 (8.9%) interventions aimed to substitute face-to-face consultations. The predominant mode of PROM administration was over the phone, followed by electronic devices and apps. Endpoints included disease activity, health care use, mortality, mental well-being, quality of life, self-efficacy, self-care, daily functioning, and other outcomes. Six studies with a low risk of bias demonstrated a positive effect or noninferiority of the PROM intervention. Conclusions: The evidence base for clinical use of PROMs is sparse, with few studies evaluated to have a low or a medium risk of bias. The clinical use of PROMs does not appear superior to usual care in the five included chronic diseases on any endpoint. To guide further research, we highlighted 6 (10.7%) studies with a low risk of bias and PROM interventions with a positive effect. These were characterized by symptom assessment with predefined cutoffs used for decision and dialogue support. Trial Registration: PROSPERO CRD42021226896; https://www.crd.york.ac.uk/PROSPERO/view/CRD42021226896

  • JMIR Med Inform 2025;13(1):e66160

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.

• 

  Machine Learning

  Sep 16, 2025

  Real-Time Estimation of Arterial Partial Pressure of Carbon Dioxide in Patients Undergoing General Anesthesia: Predictive Modeling Study

  View Authors View Abstract

  Authors List:

  • Ah Ra Lee  
  • Jun Ho Lee  
  • Sooyoung Yoo  
  • Ho-Young Lee  
  • Hyun Ho Kim  

  Abstract:

  Background: Adequate ventilation in mechanically ventilated patients during general anesthesia is contingent upon the monitoring of the arterial partial pressure of carbon dioxide (PaCO2). Despite its significance, continuous monitoring remains challenging due to the limitations of intermittent invasive measurements, such as arterial blood gas analysis (ABGA), and the imprecision of estimates and often unpredictable gradient between PaCO2 and ETCO2 may lead to misinterpretation of a patient’s ventilator status, potentially resulting in adverse outcomes. Objective: This study aimed to develop a machine learning model to continuously estimate PaCO2 in mechanically ventilated patients using a comprehensive set of readily available noninvasive parameters, thereby improving intraoperative monitoring accuracy under general anesthesia. Methods: This retrospective study used the VitalDB dataset, a public database from Seoul National University Hospital, containing records from 6,388 noncardiac surgery patients between August 2016 and June 2017. After applying inclusion and exclusion criteria, data from 2,304 surgical cases, yielding 4,651 PaCO2 measurement event points, were included in this analysis. The CatBoostRegressor model was employed to predict PaCO2. A total of 19 noninvasive features were used, comprising intraoperative vital signs such as ETCO2, body temperature, the SpO2/FiO2 ratio, respiratory rate, and airway pressures, along with preoperative clinical information including age, gender, and pulmonary function test results. The model’s performance was evaluated using a nested cross-validation scheme to ensure robust and generalizable results. Performance was assessed across hypocapnic (<35 mmHg), normocapnic (35-45 mmHg), and hypercapnic (>45 mmHg) subgroups and compared to two conventional baseline methods: a simple offset (ETCO2 + 5 mmHg) and linear regression with ETCOw as the sole predictor. Results: The developed model demonstrated superior overall performance compared to both traditional estimation methods. It achieved a mean absolute error (MAE) of 2.38 mmHg and a root mean squared error (RMSE) of 3.26 mmHg. The model showed excellent agreement with actual PaCO2 measurements, with an average intraclass correlation coefficient (ICC) of 0.87 (95% CI: 0.86-0.87). In terms of clinical utility, 90.02% of the model’s estimations fell within the highly acceptable range of ±5 mmHg error, a substantial improvement from the 80.43% achieved by the linear regression baseline. Furthermore, clinically unacceptable errors (> ±10 mmHg) were reduced to 1.20%, less than half the rate of the baseline model (2.64%). These performance improvements were consistently observed across all PaCO2 subgroups, including the more challenging hypocapnic and hypercapnic ranges. Conclusions: The developed machine learning-based model provides more accurate and reliable estimates of PaCO2 than traditional ETCO2-based methods. This approach shows potential for enhancing continuous respiratory monitoring, facilitating timely and precise clinical interventions, and serving as a valuable supplementary tool for anesthetic management. Further validation, including prospective studies to assess its impact on clinical decision-making and patient outcomes, is necessary to fully realize its clinical integration.

  • JMIR Med Inform 2025;13(1):e64855

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.

• 

  Machine Learning

  Sep 16, 2025

  Interpretable Machine Learning for Predicting Adverse Pregnancy Outcomes in Gestational Diabetes: Retrospective Cohort Study

  View Authors View Abstract

  Authors List:

  • Jiaxi Li  
  • Xiali Liu  
  • Shenyang He  
  • Yan Ren  

  Abstract:

  Background: Gestational diabetes mellitus (GDM) affects over 5% of pregnancies globally, elevating risks of type 2 diabetes postpartum and complications such as fetal death, miscarriage, and congenital abnormalities. Effective GDM management is essential to balance glycemic control and pregnancy outcomes. Objective: To develop interpretable machine learning models using GDM datasets for predicting adverse pregnancy outcomes and identifying key factors through the SHAP algorithm, thus supporting improved maternal and infant health. Methods: Data preprocessing and feature selection were performed, with ADASYN used to address class imbalance. Classification models, including logistic regression, random forest, SVM, and XGBoost, were built and enhanced through stacking method. Model interpretability was assessed with SHAP to quantify feature contributions. Results: Among 1,670 patients, 200 experienced adverse outcomes. The stacked model achieved 97.9% accuracy and an AUC of 0.96 in external validation. SHAP analysis highlighted key predictive factors, such as gestational age, fasting glucose, and blood pressure, supporting model reliability. Conclusions: This study underscores the potential of machine learning in predicting adverse outcomes in GDM, with interpretable features offering valuable clinical insights to enhance pregnancy management and maternal-infant health.

  • JMIR Med Inform 2025;13(1):e71539

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.

• 

  eHealth Infrastructures

  Sep 15, 2025

  Assessing Internet Quality Across Public Health Centers in Indonesia: Cross-Sectional Evaluation Study

  View Authors View Abstract

  Authors List:

  • Dewi Nur Aisyah  
  • Agus Heri Setiawan  
  • Alfiano Fawwaz Lokopessy  
  • Chyntia Aryanti Mayadewi  
  • M Thoriqul Aziz Endryantoro  
  • Viktor Wibowo  
  • Sarah Disviana  
  • Indra Laksana  
  • Mohammad Aviandito  
  • Zisis Kozlakidis  
  • Logan Manikam  

  Abstract:

  Background: Primary health care centers (Puskesmas) serve as the cornerstone of Indonesia’s healthcare system, providing integrated services aimed at improving individual health through prevention, treatment, and health promotion. To fulfill these roles effectively, robust technological infrastructure—particularly reliable internet connectivity—is increasingly essential. Assessing the availability and quality of internet access in Puskesmas is therefore a critical step in understanding their readiness to implement digital health initiatives and fulfill their responsibilities in delivering accessible and effective healthcare services. Objective: This study provides a national baseline assessment of internet quality and its relevant information technology infrastructure in over 10,000 Puskesmas across Indonesia. Methods: A cross-sectional survey was taken throughout all Puskesmas (10,382) in 34 provinces in Indonesia, using an online questionnaire. Categorization was done to analyze internet quality level results. Results: A total of 10,378 public health centres (99.96%) participated in this study, with 745 (7.18%) did not have internet access, 1,487 (14,33%) have limited internet access, 5,567 (53.64%) have sufficient internet access, and 2,579 (24.85%) have sufficient and fast internet access. Moreover, 832 Puskesmas (8.02%) do not have 24-hour electricity, 44,196 (43.7%) have CPU with i3 specifications, 43,044 (42.56%) have 512 GB hard disk capacity, and 67,272 (66,5%) uses antivirus. Conclusions: Although 79% of Puskesmas in Indonesia already had sufficient internet access, 21% still have limited and insufficient access. To ensure universal internet availability, it is essential to build collaborative support among internet providers and government to foster the availability and utilisation of internet satellites, high-quality computers, and electrical power to support internet connectivity.

  • JMIR Med Inform 2025;13(1):e65940

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.

• 

  Implementation Report

  Sep 12, 2025

  Lessons Learned From Building a Data Platform for Longitudinal, Analytical Use Cases and Scaling to 77 German Hospitals: Implementation Report

  View Authors View Abstract

  Authors List:

  • Markus Bockhacker  
  • Peter Martens  
  • Clara von Münchow  
  • Sarah Löser  
  • Rosita Günther  
  • Ralf Kuhlen  
  • Olaf Kannt  
  • Sebastian Ortleb  

  Abstract:

  Background: Increasing adoption of electronic medical records (EHR) enables research on real-world data. In Germany this has been limited to university hospitals, data from acute care hospitals below university level is lacking. To address this gap we initiated the Helios Safe Medical Data-Platform (HeSaMeDa), which aggregates and standardizes pseudonymised EHR data with patients’ consent. Objective: To report on the design, implementation, patient participation and lessons learned during the scaling of a research platform to incorporate consented real-world data from 77 distinct hospitals into a unified data lake. Methods: Due to variations in EHR adoption, IT infrastructure, software vendors, interface availability and regulatory requirements, we used an agile development cycle that involves constant, incremental standardization of data. We implemented a layered lambda infrastructure built on Apache Hadoop. Decentralized connectors ensure data minimization and pseudonymization. Results: We successfully scaled our data model both laterally and horizontally in 77 hospitals. However, we encountered issues during the scaling of real-time data pipelines and IHE interfaces. During the first 2 years patients were asked to consent to secondary data use 1,475,244 times during inpatient admission. We registered 1,023,633 broad consents (consent rate: 70.2%). Conclusions: Patients are generally willing to provide consent for secondary use of their data, but obtaining consent requires considerable effort. Building a research data platform isn’t an end goal, but rather a necessary step in collecting and standardizing longitudinal data to enable research on real-world data. Through the combination of agile development, phased rollouts and very high levels of automation, we have been able to achieve fast turnaround times for incorporating user feedback and are constantly improving data quality and standardization.

  • JMIR Med Inform 2025;13(1):e69853

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.

• 

  Reviews in Medical Informatics

  Sep 12, 2025

  Performance of Natural Language Processing for Information Extraction From Electronic Health Records Within Cancer: Systematic Review

  View Authors View Abstract

  Authors List:

  • Simon Dahl  
  • Martin Bøgsted  
  • Tomer Sagi  
  • Charles Vesteghem  

  Abstract:

  Background: Over the last decade, natural language processing (NLP) has provided various solutions for information extraction (IE) from textual clinical data. In recent years, the use of NLP in cancer research has gained considerable attention, with numerous studies exploring the effectiveness of various NLP techniques for identifying and extracting cancer-related entities from clinical text data. Objective: We aimed to summarize the performance differences between various NLP models for IE within the context of cancer to provide an overview of the relative performance of existing models. Methods: This systematic literature review was conducted using three databases (PubMed, Scopus, and Web of Science) to search for articles extracting cancer-related entities from clinical texts. 33 articles were eligible for inclusion. We extracted NLP models and their performance by F1 scores. Each model was categorized into the following categories: Rule-based, Traditional Machine Learning, CRF-based, Neural Network, and Bidirectional transformer. The average of the performance difference for each combination of categorizations was calculated across all articles. Results: The articles covered various scenarios, with the best performance for each article, ranging from 0.355 to 0.985 in F1 score. Looking at the overall relative performances, the bidirectional transformer category outperformed every other category (by between 0.2335 and 0.0439 on average F1 score). The percentage of articles on implementing bidirectional transformers has increased over the years. Conclusions: NLP has demonstrated the ability to identify and extract cancer-related entities from unstructured textual data. Generally, more advanced models outperform less advanced ones. The bidirectional transformer category performed the best.

  • JMIR Med Inform 2025;13(1):e68707

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.

• 

  Methods and Instruments in Medical...

  Sep 12, 2025

  Extracting Symptoms of Complex Conditions From Online Discourse (Subreddit to Symptomatology): Lexicon-Based Approach

  View Authors View Abstract

  Authors List:

  • Bushra Hossain  
  • Sarah M Preum  
  • Md Fazle Rabbi  
  • Rifat Ara  
  • Mohammed Eunus Ali  

  Abstract:

  Background: Millions of people affected with complex medical conditions with diverse symptoms often turn to online discourse to share their experiences. While some studies have explored natural language processing methods and medical information extraction tools, these typically focus on generic symptoms in clinical notes and struggle to identify patient-reported, disease-specific, subtle symptoms from online health discourse. Objective: We aimed to extract patient-reported, disease-specific symptoms shared on social media reflecting the lived experiences of thousands of affected individuals and explore the characteristics, prevalence, and occurrence patterns of the symptoms. Methods: We propose a lexicon-based symptom extraction (LSE) method to identify a diverse list of disease-specific, patient-reported symptoms. We initially used a large language model to accelerate the extraction of symptom-related key phrases that formed the lexicon. We evaluated the effectiveness of lexicon extraction against human annotation using a Jaccard index score. We then leveraged BERT-Base, BioBERT, and Phrase-BERT–based embeddings to learn representations of these symptom-related key phrases and cluster similar symptoms using k-means and hierarchical density-based spatial clustering of applications with noise (HDBSCAN). Among the different options explored in our experiments, BioBERT-based k-means clustering was found to be the most effective. Finally, we applied symptom normalization to eliminate duplicate and redundant entries in the comprehensive symptom list. Results: In a real-world polycystic ovary syndrome (PCOS) subreddit dataset, we found that LSE significantly outperformed state-of-the-art baselines, achieving at least 41% and 20% higher F₁-scores (mean 86.10) than automatic medical extraction tools and large language models, respectively. Notably, the comprehensive list of 64 PCOS symptoms generated via LSE ensured extensive coverage of symptoms reported in 7 reputable eHealth forums. Analyzing PCOS symptomatology revealed 28 potentially emerging symptoms and 8 self-reported comorbidities co-occurring with PCOS. Conclusions: The comprehensive patient-reported, disease-specific symptom list can help patients and health practitioners resolve uncertainties surrounding the disease, eliminating the variability of PCOS symptoms prevailing in the community. Analyzing PCOS symptomatology across varied dimensions provides valuable insights for public health research.

  • JMIR Med Inform 2025;13(1):e70940

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.

• 

  Computer-Aided Diagnosis

  Sep 11, 2025

  Enhancing Oral Health Diagnostics With Hyperspectral Imaging and Computer Vision: Clinical Dataset Study

  View Authors View Abstract

  Authors List:

  • Paul Römer  
  • Jean-Jacques Ponciano  
  • Katharina Kloster  
  • Fabia Siegberg  
  • Bastian Plaß  
  • Shankeeth Vinayahalingam  
  • Bilal Al-Nawas  
  • Peer W Kämmerer  
  • Thomas Klauer  
  • Daniel Thiem  

  Abstract:

  Background: Diseases of the oral cavity, including oral squamous cell carcinoma (OSCC) pose major challenges to healthcare worldwide due to its late diagnosis and complicated differentiation of oral tissues. Endoscopic hyperspectral imaging (eHSI) represents a promising approach to the demand for modern, non-invasive tissue diagnostics. This dataset is designed to enhance the performance of deep learning models by providing comprehensive spectral data essential for distinguishing between healthy and pathological oral tissue conditions. Objective: To develop and validate a clinical dataset of endoscopic hyperspectral imaging (eHSI) of the oral cavity and to evaluate the performance of deep learning-based semantic segmentation models for automated tissue classification. Methods: This clinical study included 226 participants (166 women, 60 men, aged 24-87). eHSI data were collected using an endoscopic hyperspectral sensor system, capturing spectral data in the range of 500-1000 nm. Each participant underwent five standardized intraoral hyperspectral scans of the cheek, palate, tongue, and teeth. RGB and eHSI images were archived in NPY format for Python analysis. Oral structures were annotated using RectLabel Pro©. DeepLabv3 with a ResNet-50 backbone was adapted for eHSI segmentation by modifying the first convolutional layer. The model was trained for 50 epochs on 70% of the dataset, with 30% for evaluation. Performance metrics (Precision, Recall, F1-score) confirmed its efficacy in distinguishing oral tissue types. Results: Preliminary analysis revealed that the Coefficient of Variation exceeded 15% for most spectral bands, indicating high variability in spectral signatures. DeepLabv3 (ResNet-101) achieved strong segmentation performance (F1-score: 0.856, Precision: 0.8506, Recall: 0.8634), excelling in key structures—mucosa (F1: 0.915), retractor (F1: 0.940), tooth (F1: 0.902), and palate (F1: 0.900). Moderate results for gingiva (F1: 0.760) and lip (F1: 0.706) suggest potential for further refinement. To assess dataset robustness, additional models were tested, including DeepLabv3 (ResNet-50), FCN (ResNet-50/101), PSPNet (ResNet-50/VGG16), and U-Net (EfficientNet-B0/ResNet-50). DeepLabv3 (ResNet-101) and U-Net (EfficientNet-B0) emerged as top performers, particularly in retractor, mucosa, and tooth segmentation. Conclusions: Conducting an in-depth hyperspectral analysis of oral tissue has facilitated the development of robust deep learning algorithms, thereby enhancing diagnostic accuracy and clinical applicability. This study integrates advanced imaging technologies with deep learning, offering significant potential to enhance non-invasive detection, classification, and therefore individualized treatment of oral diseases. Clinical Trial: Deutsche Forschungsgemeinschaft (DFG) - Projectnumber 516210826 https://gepris.dfg.de/gepris/projekt/516210826

  • JMIR Med Inform 2025;13(1):e76148

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.

• 

  Decision Support for Health...

  Sep 10, 2025

  Using Machine Learning to Predict-Then-Optimize Elective Orthopedic Surgery Scheduling to Improve Operating Room Utilization: Retrospective Study

  View Authors View Abstract

  Authors List:

  • Johnathan R Lex  
  • Aazad Abbas  
  • Jacob Mosseri  
  • Jay Singh Toor  
  • Michael Simone  
  • Bheeshma Ravi  
  • Cari Whyne  
  • Elias B Khalil  

  Abstract:

  Background: Total knee and hip arthroplasty (TKA and THA) are among the most performed elective procedures. Rising demand and the resources intensive nature of these procedures has contributed to longer wait times despite significant healthcare investment. Current scheduling methods often rely on average surgical durations, overlooking patient-specific variability. Objective: To determine the potential for improving elective surgery scheduling for total knee and hip arthroplasty (TKA and THA, respectively) by utilizing a two-stage approach that incorporates machine learning (ML) prediction of the duration of surgery (DOS) with scheduling optimization. Methods: Two ML models (one each for TKA and THA) were trained to predict DOS using patient factors based on 302,490 and 196,942 patients, respectively, from a large international database. Three optimization formulations based on varying surgeon flexibility were compared: Any (surgeons could operate in any operating room at any time), Split (limitation of two surgeons per operating room per day) and MSSP (limit of one surgeon per operating room per day). Two years of daily scheduling simulations were performed for each optimization problem using ML-prediction or mean DOS over a range of schedule parameters. Constraints and resources were based on a high-volume arthroplasty hospital in Canada. Results: The TKA and THA prediction models achieved test accuracy (with a 30-minute buffer) of 78.1% (MSE 0.898) and 75.4% (MSE 0.916), respectively. Any scheduling formulation performed significantly worse than the Split and MSSP formulations with respect to overtime and underutilization (P<.001). The latter two problems performed similarly (P>.05) over most schedule parameters. The ML-prediction schedules outperformed those generated using a mean DOS for most scheduling parameters, with overtime reduced on average by 300 to 500 minutes per week (12-20 minutes per operating room per day) (P <.001). However, there was more OR underutilization with the ML-prediction schedules, with it ranging from 70-192 minutes more underutilization (P<.001). Using a 15-minute schedule granularity with a waitlist pool of minimum one month generated the ML-schedule that outperformed the mean schedule 97.1% of times. Conclusions: Assuming a full waiting list, optimizing an individual surgeon’s elective operating room time using an ML-assisted predict-then-optimize scheduling system improves overall operating room efficiency, significantly decreasing overtime. This has significant potential implications for healthcare systems struggling with pressures of rising costs and growing operative waitlists.

  • JMIR Med Inform 2025;13(1):e70857

  • Choose one of the following formats:

  • Share this article:

    Or, copy the article link to your clipboard:

    Article URL

  • Citation

    • END

      for: Endnote

    • BibTeX

      for: BibDesk and LaTeX

    • RIS

      for: RefMan, Procite, Endnote, RefWorks

    Add this article to your Mendeley library Add this article to your CiteULike library Copy Citation to Clipboard

  • Metrics

    Discover Social Media mentions by hovering over the donut. Click the 'details' link for a full report.