A community effort to assess and improve computerized interpretation of 12-lead resting electrocardiogram.
Ontology highlight
ABSTRACT:
SUBMITTER: Ding Z
PROVIDER: S-EPMC8724189 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:BackgroundAccurate interpretation of a 12-lead electrocardiogram (ECG) demands high levels of skill and expertise. Early training in medical school plays an important role in building the ECG interpretation skill. Thus, understanding how medical students perform the task of interpretation is important for improving this skill.ObjectiveWe aimed to use eye tracking as a tool to research how eye fixation can be used to gain a deeper understanding of how medical students interpret ECGs.MethodsIn total, 16 medical students were recruited to interpret 10 different ECGs each. Their eye movements were recorded using an eye tracker. Fixation heatmaps of where the students looked were generated from the collected data set. Statistical analysis was conducted on the fixation count and duration using the Mann-Whitney U test and the Kruskal-Wallis test.ResultsThe average percentage of correct interpretations was 55.63%, with an SD of 4.63%. After analyzing the average fixation duration, we found that medical students study the three lower leads (rhythm strips) the most using a top-down approach: lead II (mean=2727 ms, SD=456), followed by leads V1 (mean=1476 ms, SD=320) and V5 (mean=1301 ms, SD=236). We also found that medical students develop a personal system of interpretation that adapts to the nature and complexity of the diagnosis. In addition, we found that medical students consider some leads as their guiding point toward finding a hint leading to the correct interpretation.ConclusionsThe use of eye tracking successfully provides a quantitative explanation of how medical students learn to interpret a 12-lead ECG.
| S-EPMC8554676 | biostudies-literature
Project description:ImportanceMillions of clinicians rely daily on automated preliminary electrocardiogram (ECG) interpretation. Critical comparisons of machine learning-based automated analysis against clinically accepted standards of care are lacking.ObjectiveTo use readily available 12-lead ECG data to train and apply an explainability technique to a convolutional neural network (CNN) that achieves high performance against clinical standards of care.Design, setting, and participantsThis cross-sectional study was conducted using data from January 1, 2003, to December 31, 2018. Data were obtained in a commonly available 12-lead ECG format from a single-center tertiary care institution. All patients aged 18 years or older who received ECGs at the University of California, San Francisco, were included, yielding a total of 365 009 patients. Data were analyzed from January 1, 2019, to March 2, 2021.ExposuresA CNN was trained to predict the presence of 38 diagnostic classes in 5 categories from 12-lead ECG data. A CNN explainability technique called LIME (Linear Interpretable Model-Agnostic Explanations) was used to visualize ECG segments contributing to CNN diagnoses.Main outcomes and measuresArea under the receiver operating characteristic curve (AUC), sensitivity, and specificity were calculated for the CNN in the holdout test data set against cardiologist clinical diagnoses. For a second validation, 3 electrophysiologists provided consensus committee diagnoses against which the CNN, cardiologist clinical diagnosis, and MUSE (GE Healthcare) automated analysis performance was compared using the F1 score; AUC, sensitivity, and specificity were also calculated for the CNN against the consensus committee.ResultsA total of 992 748 ECGs from 365 009 adult patients (mean [SD] age, 56.2 [17.6] years; 183 600 women [50.3%]; and 175 277 White patients [48.0%]) were included in the analysis. In 91 440 test data set ECGs, the CNN demonstrated an AUC of at least 0.960 for 32 of 38 classes (84.2%). Against the consensus committee diagnoses, the CNN had higher frequency-weighted mean F1 scores than both cardiologists and MUSE in all 5 categories (CNN frequency-weighted F1 score for rhythm, 0.812; conduction, 0.729; chamber diagnosis, 0.598; infarct, 0.674; and other diagnosis, 0.875). For 32 of 38 classes (84.2%), the CNN had AUCs of at least 0.910 and demonstrated comparable F1 scores and higher sensitivity than cardiologists, except for atrial fibrillation (CNN F1 score, 0.847 vs cardiologist F1 score, 0.881), junctional rhythm (0.526 vs 0.727), premature ventricular complex (0.786 vs 0.800), and Wolff-Parkinson-White (0.800 vs 0.842). Compared with MUSE, the CNN had higher F1 scores for all classes except supraventricular tachycardia (CNN F1 score, 0.696 vs MUSE F1 score, 0.714). The LIME technique highlighted physiologically relevant ECG segments.Conclusions and relevanceThe results of this cross-sectional study suggest that readily available ECG data can be used to train a CNN algorithm to achieve comparable performance to clinical cardiologists and exceed the performance of MUSE automated analysis for most diagnoses, with some exceptions. The LIME explainability technique applied to CNNs highlights physiologically relevant ECG segments that contribute to the CNN's diagnoses.
| S-EPMC8340011 | biostudies-literature
Project description:In clinical trials, traditionally only a limited number of 12-lead resting electrocardiograms (ECGs) can be recorded and, thus, long intervals may elapse between assessment timepoints and valuable information may be missed during times when patients' cardiac electrical activity is not being monitored. These limitations have led to the increasing use of Holter recorders which provide continuous data registrations while reducing the burden on patients and freeing up time for clinical trial staff to perform other tasks. However, there is a shortage of data comparing the two approaches. In this study, data from a randomized, double-blind, four-period, crossover thorough QT study in 40 healthy subjects were used to compare continuous 12-lead Holter recordings to standard 12-lead resting ECGs which were recorded in parallel. Heart rate and QT interval data were estimated by averaging three consecutive heartbeats. Values exceeding the sample average by more than 5% were tagged as outliers and excluded from the analysis. Visual comparisons of the ECG waveforms of the Holter signal showed a good correlation with resting ECGs at matching timepoints. Resting ECG data revealed sex differences that Holter data did not show. Specifically, women were found to have a longer QTcF of 20 ms, while men had a lower heart rate. We found that continuous recordings provided a more accurate reflection of changes in cardiac electrical activity over 24 hr. However, manual adjudication is still required to ensure the quality and accuracy of ECG data, and that only artifacts are removed thereby avoiding loss of true signals.
| S-EPMC8411681 | biostudies-literature
Project description:Predicting the best treatment strategy from genomic information is a core goal of precision medicine. Here we focus on predicting drug response based on a cohort of genomic, epigenomic and proteomic profiling data sets measured in human breast cancer cell lines. Through a collaborative effort between the National Cancer Institute (NCI) and the Dialogue on Reverse Engineering Assessment and Methods (DREAM) project, we analyzed a total of 44 drug sensitivity prediction algorithms. The top-performing approaches modeled nonlinear relationships and incorporated biological pathway information. We found that gene expression microarrays consistently provided the best predictive power of the individual profiling data sets; however, performance was increased by including multiple, independent data sets. We discuss the innovations underlying the top-performing methodology, Bayesian multitask MKL, and we provide detailed descriptions of all methods. This study establishes benchmarks for drug sensitivity prediction and identifies approaches that can be leveraged for the development of new methods.
| S-EPMC4547623 | biostudies-literature
Project description:ObjectivesElectrocardiogram (ECG) data are important for the study of cardiovascular disease and adverse drug reactions. Although the development of analytical techniques such as machine learning has improved our ability to extract useful information from ECGs, there is a lack of easily available ECG data for research purposes. We previously published an article on a database of ECG parameters and related clinical data (ECG-ViEW), which we have now updated with additional 12-lead waveform information.MethodsAll ECGs stored in portable document format (PDF) were collected from a tertiary teaching hospital in Korea over a 23-year study period. We developed software which can extract all ECG parameters and waveform information from the ECG reports in PDF format and stored it in a database (meta data) and a text file (raw waveform).ResultsOur database includes all parameters (ventricular rate, PR interval, QRS duration, QT/QTc interval, P-R-T axes, and interpretations) and 12-lead waveforms (for leads I, II, III, aVR, aVL, aVF, V1, V2, V3, V4, V5, and V6) from 1,039,550 ECGs (from 447,445 patients). Demographics, drug exposure data, diagnosis history, and laboratory test results (serum calcium, magnesium, and potassium levels) were also extracted from electronic medical records and linked to the ECG information.ConclusionsElectrocardiogram information that includes 12 lead waveforms was extracted and transformed into a form that can be analyzed. The description and programming codes in this case report could be a reference for other researchers to build ECG databases using their own local ECG repository.
| S-EPMC6085199 | biostudies-literature
Project description:BackgroundResting 12-lead electrocardiography is widely used for the detection of cardiac diseases. Electrocardiogram readings have been reported to be affected by aging and, therefore, can predict patient mortality.MethodsA total of 12,837 patients without structural heart disease who underwent electrocardiography at baseline were identified in the Shinken Database among those registered between 2010 and 2017 (n = 19,170). Using 438 electrocardiography parameters, predictive models for all-cause death and cardiovascular (CV) death were developed by a support vector machine (SVM) algorithm.ResultsDuring the observation period of 320.4 days, 55 all-cause deaths and 23 CV deaths were observed. In the SVM prediction model, the mean c-statistics of 10 cross-validation models with training and testing datasets were 0.881 ± 0.027 and 0.927 ± 0.101, respectively, for all-cause death and 0.862 ± 0.029 and 0.897 ± 0.069, respectively for CV death. For both all-cause and CV death, high values of permutation importance in the ECG parameters were concentrated in the QRS complex and ST-T segment.ConclusionsParameters acquired from 12-lead resting electrocardiography could be applied to predict the all-cause and CV deaths of patients without structural heart disease. The ECG parameters that greatly contributed to the prediction were concentrated in the QRS complex and ST-T segment.
| S-EPMC7874456 | biostudies-literature
Project description:Electrocardiogram (ECG) is a widely used reliable, non-invasive approach for cardiovascular disease diagnosis. With the rapid growth of ECG examinations and the insufficiency of cardiologists, accurate and automatic diagnosis of ECG signals has become a hot research topic. In this paper, we developed a deep neural network for automatic classification of cardiac arrhythmias from 12-lead ECG recordings. Experiments on a public 12-lead ECG dataset showed the effectiveness of our method. The proposed model achieved an average F1 score of 0.813. The deep model showed superior performance than 4 machine learning methods learned from extracted expert features. Besides, the deep models trained on single-lead ECGs produce lower performance than using all 12 leads simultaneously. The best-performing leads are lead I, aVR, and V5 among 12 leads. Finally, we employed the SHapley Additive exPlanations method to interpret the model's behavior at both the patient level and population level.
| S-EPMC8082080 | biostudies-literature
Project description:Abstract Background We have demonstrated that a neural network is able to predict a person’s age from the electrocardiogram (ECG) [artificial intelligence (AI) ECG age]. However, some discrepancies were observed between ECG-derived and chronological ages. We assessed whether the difference between AI ECG and chronological age (Age-Gap) represents biological ageing and predicts long-term outcomes. Methods and results We previously developed a convolutional neural network to predict chronological age from ECGs. In this study, we used the network to analyse standard digital 12-lead ECGs in a cohort of 25 144 subjects ≥30 years who had primary care outpatient visits from 1997 to 2003. Subjects with coronary artery disease, stroke, and atrial fibrillation were excluded. We tested whether Age-Gap was correlated with total and cardiovascular mortality. Of 25 144 subjects tested (54% females, 95% Caucasian) followed for 12.4 ± 5.3 years, the mean chronological age was 53.7 ± 11.6 years and ECG-derived age was 54.6 ± 11 years (R2 = 0.79, P < 0.0001). The mean Age-Gap was small at 0.88 ± 7.4 years. Compared to those whose ECG-derived age was within 1 standard deviation (SD) of their chronological age, patients with Age-Gap ≥1 SD had higher all-cause and cardiovascular disease (CVD) mortality. Conversely, subjects whose Age-Gap was ≤1 SD had lower all-cause and CVD mortality. Results were unchanged after adjusting for CVD risk factors and other survival influencing factors. Conclusion The difference between AI ECG and chronological age is an independent predictor of all-cause and cardiovascular mortality. Discrepancies between these possibly reflect disease independent biological ageing. Graphical Abstract In subjects without baseline cardiovascular disease, an older artificial intelligence (AI)-derived electrocardiogram (ECG) Age (AI-ECG age) compared to chronological age was associated with increased all-cause and cardiovascular mortality. Subjects, where the reverse was true, had lower mortality. Discrepancies between chronological age and ECG-derived age possibly reflect disease independent biological ageing.
| S-EPMC9707884 | biostudies-literature
Project description:Various efforts have been made to diagnose acute cardiovascular diseases (CVDs) early in patients. However, the sole option currently is symptom education. It may be possible for the patient to obtain an early 12-lead electrocardiogram (ECG) before the first medical contact (FMC), which could decrease the physical contact between patients and medical staff. Thus, we aimed to verify whether laypersons can obtain a 12-lead ECG in an off-site setting for clinical treatment and diagnosis using a patch-type wireless 12-lead ECG (PWECG). Participants who were ≥ 19 years old and under outpatient cardiology treatment were enrolled in this simulation-based one-arm interventional study. We confirmed that participants, regardless of age and education level, can use the PWECG on their own. The median age of the participants was 59 years (interquartile range [IQR] = 56-62 years), and the median duration to obtain a 12-lead ECG result was 179 s (IQR = 148-221 s). With appropriate education and guidance, it is possible for a layperson to obtain a 12-lead ECG, minimizing the contact with a healthcare provider. These results can be used subsequently for treatment.
| S-EPMC10004446 | biostudies-literature
Project description:BackgroundHandheld single-lead electrocardiograms (1L-ECG) present a welcome addition to the diagnostic arsenal of general practitioners (GPs). However, little is known about GPs' 1L-ECG interpretation skills, and thus its reliability in real-world practice.ObjectiveTo determine the diagnostic accuracy of GPs in diagnosing atrial fibrillation or flutter (AF/Afl) based on 1L-ECGs, with and without the aid of automatic algorithm interpretation, as well as other relevant ECG abnormalities.MethodsWe invited 2239 Dutch GPs for an online case-vignette study. GPs were asked to interpret four 1L-ECGs, randomly drawn from a pool of 80 case-vignettes. These vignettes were obtained from a primary care study that used smartphone-operated 1L-ECG recordings using the AliveCor KardiaMobile. Interpretation of all 1L-ECGs by a panel of cardiologists was used as reference standard.ResultsA total of 457 (20.4%) GPs responded and interpreted a total of 1613 1L-ECGs. Sensitivity and specificity for AF/Afl (prevalence 13%) were 92.5% (95% CI: 82.5-97.0%) and 89.8% (95% CI: 85.5-92.9%), respectively. PPV and NPV for AF/Afl were 45.7% (95% CI: 22.4-70.9%) and 98.8% (95% CI: 97.1-99.5%), respectively. GP interpretation skills did not improve in case-vignettes where the outcome of automatic AF-detection algorithm was provided. In detecting any relevant ECG abnormality (prevalence 22%), sensitivity, specificity, PPV and NPV were 96.3% (95% CI: 92.8-98.2%), 68.8% (95% CI: 62.4-74.6%), 43.9% (95% CI: 27.7-61.5%) and 97.9% (95% CI: 94.9-99.1%), respectively.ConclusionsGPs can safely rule out cardiac arrhythmias with 1L-ECGs. However, whenever an abnormality is suspected, confirmation by an expert-reader is warranted.
| S-EPMC8006764 | biostudies-literature