Physiological measurement最新文献

{"title":"CDG-MACE score: an interpretable scoring model for risk stratification in emergency chest pain with normal ECG.","authors":"Qinghua Sun, Chunmiao Liang, Tianyuan Qi, Rugang Liu, Jiali Li, Hao Zhang, Jiaojiao Pang, Yuguo Chen, Cong Wang","doi":"10.1088/1361-6579/ae601a","DOIUrl":"10.1088/1361-6579/ae601a","url":null,"abstract":"<p><p><i>Objective.</i>Rapid stratification of acute chest pain patients with non-ischemic electrocardiograms (ECGs) remains challenging. We developed and externally validated an interpretable cardiodynamicsgram (CDG)- major adverse cardiovascular events (MACE) Score to predict 30 d MACE.<i>Approach.</i>We proposed a three-step framework: (1) ECG dynamic analysis: using deterministic learning to model the ST-T repolarization process and derive CDG features that capture subtle repolarization abnormalities. We defined the temporal heterogeneity index and spatial heterogeneity index as quantitative CDG. (2) Ensemble model: training an XGBoost classifier on eight pre-specified variables with five-fold cross-validation. Patient-level splits were used; only the first emergency department ECG per patient entered the model. (3) Score derivation: transforming the ensemble into a sparse, globally interpretable score via SHAP-based variable contributions. To mitigate the demographic influence, age and gender were included as explicit covariates, and the study results were prespecified to be reported stratified by age and gender in both cohorts.<i>Main results.</i>Calibration and decision-analytic utility were assessed. Two independent cohorts (<i>n</i>= 2836) were included. In Cohort-1 (<i>n</i>= 2196; 23.27% MACE), the ensemble model achieved AUC 0.8441. Adding CDG dynamics to clinical variables improved discrimination compared with a clinical-only model (AUC 0.7963-0.8221). The derived CDG-MACE Score maintained discrimination (internal AUC 0.8221) and generalized well to Cohort-2 (<i>n</i>= 640; 11.09% MACE; external AUC 0.8219). Using prespecified cutoffs from the training set (low ⩽ 9.52; high > 26.83), the internal low-risk group had negative predictive value (NPV) 99.22% and MACE 0.78%, while the external low-risk group achieved NPV 100%. Ablation analyses confirmed that CDG dynamics contributed independent signals beyond demographics.<i>Significance.</i>The CDG-MACE Score combines dynamic ECG modeling with a SHAP-linearized scoring system to achieve discrimination with global interpretability, enabling safe exclusion of low-risk patients without typical ischemic ECG changes. External validation suggests robustness and clinical utility; additional multicenter prospective studies and fairness monitoring are warranted.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":" ","pages":""},"PeriodicalIF":2.7,"publicationDate":"2026-05-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147691427","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Investigating the interrelationship of pulse wave dynamics across the menstrual cycle and pregnancy.","authors":"Lin Yang, Yanting Qu, Cuiting Lian, Jiarui Wang, Langlang Ding, Fang Zhao, Mengdi Gao","doi":"10.1088/1361-6579/ae622b","DOIUrl":"10.1088/1361-6579/ae622b","url":null,"abstract":"<p><p><i>Objective.</i>This study aims to investigate the pulse wave characteristics across the menstrual cycle and pregnancy, comparing their cardiovascular adaptations to elucidate shared physiological mechanisms and potential predictive markers for pregnancy-related cardiovascular risks.<i>Approach.</i>The study analyzes pulse wave characteristics, including hemodynamic parameters (cardiac output (CO) and systemic vascular resistance), waveform features, and pulse wave characteristic indices. The analysis spans two physiological cycles: (1) the menstrual cycle, with parallel evaluations of normal and dysmenorrhea groups; and (2) stratified pregnancy stages (early, mid, and late), with comparative assessments between normal and abnormal blood parameter groups.<i>Main results.</i>Both the menstrual cycle and pregnancy exhibit analogous pulse wave variations, reflecting hemodynamic adjustments in CO and vascular elasticity. Pregnancy stages demonstrate progressive pulse wave alterations, with abnormal blood parameters correlating with distinct waveform deviations. Menstrual cycle patterns provide a foundational model for these adaptive changes.<i>Significance.</i>The findings reveal a physiological continuum between menstrual and gestational cardiovascular adaptations, highlighting pulse wave analysis as a potential tool for early risk stratification. This study advances the understanding of female cardiovascular dynamics, offering implications for targeted health monitoring and intervention strategies.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":" ","pages":""},"PeriodicalIF":2.7,"publicationDate":"2026-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147729578","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A multimodal fusion network for heart sound abnormality detection and classification.","authors":"Hong Duc Nguyen, Duc Tri Phan","doi":"10.1088/1361-6579/ae6415","DOIUrl":"10.1088/1361-6579/ae6415","url":null,"abstract":"<p><p><i>Objective.</i>Accurate physiological assessment of cardiac function from heart sounds remains challenging due to background noise, variable heart rates, and the need for reliable cardiac-cycle segmentation. This study aimed to develop a fully E2E deep learning framework that extracts diagnostic information directly from raw heart sound recordings for cardiac abnormality detection and classification.<i>Approach.</i>We propose HS-MMNet, an E2E multi-modal deep learning framework designed for physiological heart sound analysis. Recordings are preprocessed (normalization and 25-400 Hz bandpass filtering) and divided into fixed-length 2.5 s segments. A Convolution Head with multi-atrous spatial pyramid and channel-spatial attention extracts fine-grained local temporal patterns from the filtered 1-D waveform. A Transformer Head captures long-range spectro-temporal dependencies from Log-Mel spectrograms. These hypotheses are iteratively fused by a novel multi-hypothesis cross-attention module with cyclic query-key-value assignment and a hypothesis-mixing MLP, enabling rich cross-site interaction and effective suppression of noise and non-informative regions. Recording-level classification is obtained via a fully connected layer.<i>Main results.</i>On the PhysioNet/CinC Challenge 2016 dataset, HS-MMNet achieved 94.80% accuracy, 92.10% sensitivity, 96.85% specificity, 87.50% precision, and 89.74%<i>F</i>1-score, outperforming all previously reported methods. On the balanced five-class Yaseen dataset (normal, aortic stenosis, mitral regurgitation, mitral stenosis, mitral valve prolapse), it attained 99.60% macro-averaged precision, recall, and<i>F</i>1-score with only four misclassifications in 1000 recordings, establishing new state-of-the-art (SOTA) benchmarks.<i>Significance.</i>HS-MMNet represents an advance in automated physiological measurement from heart sounds. By eliminating cardiac cycle detection and multi-channel requirements while achieving SOTA diagnostic performance, it provides a practical, scalable solution for accurate cardiovascular screening with primary-care and low-resource settings.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":" ","pages":""},"PeriodicalIF":2.7,"publicationDate":"2026-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147778136","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Simultaneous macrovasculature and microvasculature cerebral autoregulation derived from the transfer function analysis of integrated cerebral haemodynamic data in older adults.","authors":"James D Ball, Ronney B Panerai, Jatinder S Minhas, Olivia Edwards, Abdulaziz Alshehri, Yvonne Sensier, Thompson Robinson, Lucy C Beishon","doi":"10.1088/1361-6579/ae63a3","DOIUrl":"10.1088/1361-6579/ae63a3","url":null,"abstract":"<p><p><i>Objective.</i>Cerebral autoregulation (CA) maintains stable cerebral blood flow (CBF) despite mean arterial pressure (MAP) fluctuations. Transcranial doppler ultrasonography (TCD) and near-infrared spectroscopy (NIRS) report CBF surrogates and facilitate CA assessment. This study aimed to investigate the distinct information they provide about dynamic CA (dCA) responses to MAP step changes<i>Approach</i>. Simultaneous TCD-NIRS was performed on 28 healthy older participants alongside continuous measurements of beat-to-beat and breath-to-breath MAP, heart rate and end-tidal CO₂. Transfer function analysis (TFA) was performed, varying input/output metrics including MAP, middle and posterior cerebral artery blood flow velocity macrovasculature velocity (MCAv/PCAv) and oxyhaemoglobin (HbO₂), comparing macro- and microvasculature dCA, respectively.<i>Main results.</i>No differences in regional MAP step change HbO₂responses were found across eight pre-frontal (<i>p</i>= 0.14) or four averaged regions (<i>p</i>= 0.69). There was a significant effect of time in HbO₂responses to MAP step change (<i>p</i>< 0.001), and to MCAv step change (<i>p</i>= 0.016). There were also significant differences between HbO₂and MCAv and PCAv responses (<i>p</i>< 0.001). Distinct TCD and NIRS step responses suggest much slower dCA responses in the microvasculature, compared to MCA and PCA, without regional differences.<i>Significance</i>. Further investigation into regional dCA differences is needed alongside potential benefits of simultaneous TCD-NIRS in pathological states.ClinicalTrials.gov ID: NCT05649800.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":" ","pages":""},"PeriodicalIF":2.7,"publicationDate":"2026-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147778234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"EdgeECG: A lightweight edge-oriented network with dual criterion pruning for real-time ECG arrhythmia classification.","authors":"Jiayan Huang, Chuansheng Wang, Antoni Grau, Yongyi Xiao, Shaoye Luo, Yan Che, Zuoyong Li","doi":"10.1088/1361-6579/ae696a","DOIUrl":"https://doi.org/10.1088/1361-6579/ae696a","url":null,"abstract":"<p><strong>Objective: </strong>Miniature electrocardiogram (ECG) devices can rapidly and accurately acquire real-time cardiac signals, enabling timely warnings for patients with heart disease. To achieve accurate arrhythmia classification on resource-constrained ECG edge devices, we propose EdgeECG, an ultra-lightweight neural network designed for deployment on low-power microcontrollers.</p><p><strong>Approach: </strong>EdgeECG is first designed with a compact convolutional architecture to ensure compatibility with resource-limited embedded platforms. A dual criterion pruning (DCP) strategy is then introduced to evaluate weight importance using both magnitude and median deviation for more precise model compression. In addition, quantization is applied to reduce storage cost and improve deployment efficiency on an STM32F103 microcontroller.</p><p><strong>Main results: </strong>Experimental results on the benchmark MIT-BIH arrhythmia dataset show that EdgeECG achieves an overall classification accuracy of 98.07%, outperforming several representative methods. In patient-independent record-level evaluation, the accuracy reaches 80.73%. The model contains only 2680 parameters and is successfully deployed on an STM32F103 microcontroller with 64 KB SRAM, achieving an inference latency of 0.025 s and an energy consumption of 0.156 mJ per inference. In addition, DCP reduces the number of non-zero parameters by nearly 50% while maintaining promising classification performance.</p><p><strong>Significance: </strong>EdgeECG provides an effective solution for five-class ECG arrhythmia classification on severely resource-constrained edge devices. Its compact architecture, effective pruning strategy, and successful deployment on an STM32F103 demonstrate its potential for low-power edge-based ECG analysis and wearable cardiac monitoring applications. The source code of EdgeECG is publicly available at: https://github.com/jyanhuang/EdgeECG-code.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":" ","pages":""},"PeriodicalIF":2.7,"publicationDate":"2026-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147841713","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Acute effects of FIFA 11+ warm-up on skin temperature in male and female amateur soccer players.","authors":"Patricia Caudet, Ernest Baiget, Abraham Batalla, Joshua Colomar, Francisco Corbi","doi":"10.1088/1361-6579/ae63a4","DOIUrl":"10.1088/1361-6579/ae63a4","url":null,"abstract":"<p><p><i>Objective</i>. Warm-up is a fundamental part of the training session and competition preparation, improving performance and reducing sports injuries. The FIFA 11+ is a specific evidence-based routine created to enhance neuromuscular performance and prevent lower-limb injuries. Infrared thermography (IRT) is a non-invasive tool for monitoring tissue state and thermoregulation responses. This study examined the acute effects of the FIFA 11+ warm-up on skin surface temperature (Tsk) patterns of the dominant lower limb in amateur football players using IRT.<i>Approach</i>. A pre-post observational design was applied to 120 amateur players (60 men, 60 women) before a match. Baseline and post-intervention Tsk measurements were acquired with a FLIR T540-EST camera following the TERMOINEF protocol.<i>Main results</i>. Significant post-warm-up Tsk reductions were detected in proximal muscle regions, particularly in quadriceps and adductors, with a reduction of -1.9 to -2.4 °C (ES = - 1.63 to -1.92, large) in women and -0.7 to -1.2 °C (ES = - 0.66 to -1.07, moderate) in men. Conversely, distal regions such as the anterior plantar arch showed marked Tsk increases of +2.6 °C (ES = 1.83, large) in women; +2.1 °C (ES = 1.42, large) in men. Men exhibited higher absolute Tsk values overall (<i>η</i>²≈ 0.17-0.26), whereas women displayed greater relative percentage changes, including sex-specific Achilles tendon response (a decrease in women versus a slight increase in men).<i>Significance</i>. FIFA 11+ induces heterogeneous, region- and sex- dependent thermal adaptations, supporting the use of IRT as a valid tool for individualized warm-up monitoring and optimization in football.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":" ","pages":""},"PeriodicalIF":2.7,"publicationDate":"2026-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147778088","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine learning-based information flow analysis of ECG signals for long QT syndrome.","authors":"Mateusz Ozimek, Małgorzata Andrzejewska-Ozimek, Monika Petelczyc","doi":"10.1088/1361-6579/ae6969","DOIUrl":"https://doi.org/10.1088/1361-6579/ae6969","url":null,"abstract":"<p><strong>Objective: </strong>Cardiovascular diseases remain the leading cause of death worldwide, highlighting the need for non-invasive and cost-effective risk assessment tools. Biological systems, including the heart, exhibit complex nonlinear dynamics arising from interactions between their subsystems. Information-theoretic measures, particularly entropy-based methods, provide a framework to quantify these interactions. Using ECG recordings, we investigate information flow between heart rhythm and ventricular repolarization to identify potential markers of pathological alterations in cardiac electrical activity. &#xD;Approach: Entropy-based measures of information transfer were derived from beat-to-beat ECG time series using a window-based approach and subsequently averaged at the subject level. These features were used as inputs to supervised machine learning models to discriminate patients with congenital long QT syndrome from healthy controls. Model performance was evaluated using repeated stratified train-test splits, and classification robustness was assessed across multiple runs using standard performance metrics, including the area under the receiver operating characteristic curve. The explainable artificial intelligence techniques were applied. SHapley Additive exPlanations (SHAP) were used to quantify the contribution of entropy-based features to the model predictions. This post-hoc explainability analysis enabled systematic assessment of feature importance while preserving the predictive performance of the models.&#xD;Results: The proposed approach achieved high and stable classification performance across repeated validation runs. Both Random Forest (RF) and Support Vector Machine (SVM) classifiers demonstrated high discrimination between long QT syndrome patients and healthy controls, with consistently high AUC. For RF a mean accuracy of 95.9%, mean sensitivity of 95.9%, and mean specificity of 92.9% were achieved across repeated runs. For SVM the corresponding mean values were 93.1%, 93.1%, and 92.0%, respectively.&#xD;Conclusions: Explainability analysis revealed a dominant contribution of multivariate and conditional information flow features compared with single-source entropy measures, highlighting the relevance of joint and conditional interactions in the classification patterns.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":" ","pages":""},"PeriodicalIF":2.7,"publicationDate":"2026-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147841732","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Minimum data length required for reliable resting frequency-domain ultra-short-term heart rate variability analysis.","authors":"Xiangni Lin, Gengxing Liu, Lin Wang, Xiaomeng Zhou, Xinping Deng, Ning Ji, Jingjing Wei, Jingxin Shi, Chunjie Yuan, Rongcai Jiang, Guanglin Li, Wan-Hua Lin","doi":"10.1088/1361-6579/ae5c54","DOIUrl":"10.1088/1361-6579/ae5c54","url":null,"abstract":"<p><p><i>Objective.</i>Heart rate variability (HRV) is widely used to assess autonomic function, but the minimum data length required for reliable ultra-short-term (UST) frequency-domain analysis remains without consensus. This study aimed to (1) identify factors determining the minimum required length and (2) determine minimum length for reliable estimation.<i>Approach.</i>Simulated relatively stationary inter-beat intervals (IBIs) were used to examine spectral factors influencing minimum data length. Relatively stationary IBIs from 20 min resting electrocardiogram (ECG) recordings in eight healthy subjects were analyzed to determine practical requirements. High frequency (HF) and low frequency (LF) estimates were compared against 5 min references, and very low frequency (VLF) against 20 min references, using limits of agreement (LoA) and intraclass correlation coefficient (ICC).<i>Results.</i>Both signal properties (spectral distribution proximity to defined band edges) and spectral analysis parameters (window and segment length) are critical factors determining the minimum required length. Data lengths of ∼60 s, 100 s, and 1000 s provided coarse estimates for HF, LF, and VLF, respectively, with ICC > 0.9. More reliable estimates were achieved at ∼90 s, 250 s, and 1080 s, where the LoA remained within ±20%. Even stricter reliability was obtained at ∼200 s, 290 s, and 1180 s, where the LoA further narrowed to within ±10%. Although LF achieved ICC > 0.90 at ∼100 s, the LoA remained wide (> ± 50%), reliable LF estimation required⩾220 s, where both absolute values and ICC stabilized.<i>Significance.</i>These findings offer methodological insights into the selection of recording durations and parameters for UST frequency-domain HRV analysis in clinical, wearable, and research applications.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":" ","pages":""},"PeriodicalIF":2.7,"publicationDate":"2026-05-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147634022","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mathematical modeling of the respiratory system in swine with acute respiratory distress syndrome.","authors":"Azadeh Dashti, Tsend-Ayush Batsaikhan, Enkhbaatar Perenlei, Chris Blakeman, Richard D Branson, Jin-Oh Hahn, Michael Kinsky, Christopher G Scully, Ramin Bighamian","doi":"10.1088/1361-6579/ae674f","DOIUrl":"https://doi.org/10.1088/1361-6579/ae674f","url":null,"abstract":"<p><p>Abstract- Objective: The objective of this study is to develop a low-order mathematical model of respiratory gas exchange for design and evaluation of physiological closed-loop controlled (PCLC) mechanical ventilation and oxygenation systems, particularly under acute respiratory distress syndrome (ARDS) conditions.</p><p><strong>Approach: </strong>Experimental data from 11 swine subjects undergoing ARDS followed by hemorrhage were used to derive the mathematical model. The animals were ventilated using a PCLC system that regulated inspired oxygen fraction (FiO2), positive end-expiratory pressure (PEEP), and other ventilation parameters. The mathematical model takes metabolic carbon dioxide production rate (V ̇CO2), FiO2, and PEEP as inputs and outputs end-tidal CO2 pressure (PetCO2), arterial oxygen pressure (PaO2), and oxygen saturation (SaO2).</p><p><strong>Main results: </strong>The mathematical model accurately reproduced observed gas exchange dynamics in ARDS conditions, effectively capturing O2 and CO2 behavior in response to the controlled ventilation parameters.</p><p><strong>Significance: </strong>The present study focuses on mathematical model development and calibration using experimental data. The current results support its utility in simulating respiratory gas exchange under lung injury conditions.</p><p><strong>Significance: </strong>This low-order mathematical model may offer a promising tool for evaluating and designing PCLC mechanical ventilation and oxygenation, with potential applications in controller development and its preclinical testing.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":" ","pages":""},"PeriodicalIF":2.7,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147819445","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A teacher-student deep learning framework for enhanced clinical screening of heart failure from 12-lead electrocardiograms.","authors":"Enhan Liu, Ye Liu, Zihang Wang, Han Zhou, Xingmei Wang","doi":"10.1088/1361-6579/ae6750","DOIUrl":"https://doi.org/10.1088/1361-6579/ae6750","url":null,"abstract":"<p><strong>Objective: </strong>Identifying heart failure (HF) from electrocardiograms (ECG) is challenging due to the lack of definitive features. This study aims to develop a deep learning-based clinical decision support system, CTTSnet, for accurate and automated HF screening using ECGs.</p><p><strong>Approach: </strong>We propose a novel teacher-student framework, CTTSnet, which synergistically integrates a Transformer as the teacher and a Convolutional Neural Network (CNN) as the student. This architecture combines global context-awareness with efficient local feature extraction. The model was trained and evaluated on a large-scale, real-world dataset of 27,018 ECGs and further validated on two external public datasets to assess generalizability.</p><p><strong>Main results: </strong>CTTSnet achieved an AUROC of 0.941 on the primary clinical dataset, surpassing strong baselines. The model attained an accuracy of 87.3%, with a recall of 88.6% and specificity of 85.8%. External validation on two additional public datasets further confirmed the model's generalizability.</p><p><strong>Significance: </strong>CTTSnet provides a scalable, high-accuracy tool for HF triage from routine ECG recordings. Its potential clinical value lies in assisting early identification of patients at risk of HF, thereby helping reduce missed diagnoses and facilitating timely referral for further evaluation.</p>","PeriodicalId":20047,"journal":{"name":"Physiological measurement","volume":" ","pages":""},"PeriodicalIF":2.7,"publicationDate":"2026-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147819456","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}