Multi-Channel Lightweight Convolution Neural Network for Anterior Myocardial Infarction Detection

Electrocardiogram (ECG) is widely used in Myocardial infarction (MI) diagnosis. The automatic diagnosis of MI based on the 12-lead ECG needs to consider not only the waveform change features in multi-resolution time series, but also the spatial correlation information between the leads. To this end, this work proposed multiscale spatiotemporal feature extraction method based on Convolutional neural network (CNN) for MI automatic diagnosis. First, the 12-lead ECG is first transformed into an ECG image through wavelet decomposition and 3-dimensional space reconstruction. The MI-CNN model is then constructed to identify MI using 41368 ECG images. Finally, we develop the LL-CNN model, which is utilized only after the ECG signal is identified as an MI event by the MI-CNN model, to localize MI by employing transfer learning to overcome the limited data problem. The proposed method has achieved an accuracy of 99.51% on MI detection, and a macro-F1 of 99.14% on MI localization. Moreover, the features visualization shows that U-wave has significant diagnostic value for MI. The proposed method significantly improves the performance of MI detection and localization compared with other methods. It is promising to be used for MI monitoring and diagnosis.

The occlusion in one of the coronary arteries of the heart leads to the cardiac ailment, myocardial infarction (MI). The localization of MI based on the investigation of the morphology of the multi-lead electrocardiogram (ECG) is the initial task for the diagnosis of this ailment. In this paper, the multiscale convolutional neural network is proposed for the automated localization of MI ailment from multi-lead electrocardiogram (ECG) beats. The Fourier-Bessel (FB) series expansion based empirical wavelet transform (EWT) with fixed order ranges is introduced for the multiscale analysis of multi-lead ECG beat. The FB spectrum of each lead ECG beat is segregated into contiguous segments using the fixed order ranges. Furthermore, the order ranges from these contiguous segments are used to design an empirical wavelet filter bank for the extraction of subband signals from each lead ECG beat. The convolutional neural network (CNN) is used for the classification of various categories of MI as anterior MI (AMI), anterio-lateral MI (ALMI), anterio-septal MI (ASMI), inferior MI (IMI), inferio-lateral MI (ILMI), inferio-posterio-lateral MI (IPLMI) and normal sinus rhythm (NSR). The experimental results reveal that the lower-order range subband signal coupled with CNN attains higher average accuracy values of 99.92%, 99.34%, 99.95%, 99.95%, 99.91%, and 99.86% respectively, for AMI, ALMI, ASMI, IMI, ILMI, and IPLMI classes. The subband signal of multi-lead ECG beats with order range of [1-26] is highly affected during various categories of MI heart disease, and this band signal has higher performance as compared to the existing MI localization approaches.

The American College of Cardiology (ACC)/American Heart Association (AHA) Task Force on Practice Guidelines was formed to make recommendations regarding the diagnosis and treatment of patients with known or suspected cardiovascular disease. Coronary artery disease (CAD) is the leading cause of death in the United States. Unstable angina (UA) and the closely related condition non–ST-segment elevation myocardial infarction (NSTEMI) are very common manifestations of this disease. These life-threatening disorders are a major cause of emergency medical care and hospitalizations in the United States. In 1996, the National Center for Health Statistics reported 1 433 000 hospitalizations for UA or NSTEMI. In recognition of the importance of the management of this common entity and of the rapid advances in the management of this condition, the need to revise guidelines published by the Agency for Health Care Policy and Research (AHCPR) and the National Heart, Lung and Blood Institute in 1994 was evident. This Task Force therefore formed the current committee to develop guidelines for the management of UA and NSTEMI. The present guidelines supersede the 1994 guidelines. The customary ACC/AHA classifications I, II, and III summarize both the evidence and expert opinion and provide final recommendations for both patient evaluation and therapy: Class I: Conditions for which there is evidence and/or general agreement that a given procedure or treatment is useful and effective . Class II: Conditions for which there is conflicting evidence and/or a divergence of opinion about the usefulness/efficacy of a procedure or treatment. Class IIa: Weight of evidence/opinion is in favor of usefulness/efficacy. Class IIb: Usefulness/efficacy is less well established by evidence/opinion. Class III: Conditions for which there is evidence and/or general agreement that the procedure/treatment is not useful/effective and in some cases may be harmful. The weight of the evidence was ranked highest (A) if the data …

Millions 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. To 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. This 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. A 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. Area 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. A 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. The 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.

Setting a New Gold Standard

The influencing factors of community risk are complex. For the low accuracy of traditional prediction model, a multichannel convolutional neural network community risk prediction model is proposed by improving convolutional neural network of deep learning. First of all, in the community risk prediction model, the structure of multichannel input convolutional neural network is selected. Then, add it into the full connection layer. Subsequently, the DenseNet layer is added to establish connections between different network layers. Finally, the receptive field is improved, and the gradient disappearance is solved. Thus, the prediction accuracy of model is improved. Compared with the traditional model, the proposed multichannel convolutional neural network model has better prediction accuracy. In addition, it performs better on the three indicators, namely, correlation coefficient R , coefficient of determination R 2 , and mean square root error RMSE. Compared with the commonly used LSTM model and logic regression model, the proposed model also has certain advantages, which is more suitable for community risk prediction.

New Insights Into the Use of the 12-Lead Electrocardiogram for Diagnosing Acute Myocardial Infarction in the Emergency Department

The diagnostic values of the aVR lead or "Vereckei algorithm," and the lead II R-wave peak time (RWPT) criterion, recently devised for the differential diagnosis of wide QRS complex tachycardias (WCTs), were compared. A total of 212 WCTs (142 ventricular tachycardias [VTs], 62 supraventricular tachycardias [SVT], and eight preexcitation SVTs) from 145 patients with proven electrophysiologic diagnoses were retrospectively analyzed by seven examiners blinded to the electrophysiologic diagnoses. The overall test accuracy of the Vereckei algorithm was superior to that of the RWPT criterion (84.3% vs. 79.6%; p = 0.0003). The sensitivity of the Vereckei algorithm for VT diagnosis was greater than that of RWPT criterion (92.4% vs. 79.1%; p < 0.0001). The negative predictive value (NPV) for the Vereckei algorithm was also greater (77.8%; 95% confidence interval [CI] = 73.6% to 82.1%) than that of the RWPT criterion (61.6%; 95% CI = 57.6% to 65.6%). The specificity of the Vereckei algorithm was lower than that of the RWPT criterion (64.7% vs. 80.9%; p < 0.0001). The positive predictive value (PPV) was also lower for the Vereckei algorithm (86.4%; 95% CI = 84.4% to 88.4%) than for the RWPT criterion (90.9%; 95% CI = 89.1% to 92.8%). Incorrect diagnoses made by the Vereckei algorithm were mainly due to misdiagnosis of SVT as VT (65.7% of cases), and those made by the RWPT criterion were due to the more dangerous misdiagnosis of VT as SVT (72.5% of cases). The Vereckei algorithm was superior in overall test accuracy, sensitivity, and NPV for VT diagnosis and inferior in specificity and PPV to the RWPT criterion. All of these parameters were lower in "real life" than those reported by the original authors for each of the particular electrocardiographic methods.

Perhaps the most useful means of monitoring the critically ill patient is the electrocardiogram (ECG). A 12lead ECG is useful to rule out cardiac complications in the acutely injured patient, the postoperative patient, and the septic patient [1]. Limb lead II or chest lead V may be continuously monitored for evidence of progressive cardiac disease and complicating dysrhythmias [1]. This section summarizes the usefulness and indications of the ECG in monitoring of the high-risk, critically ill patient and possibilities for future clinical use. Interpretation of the ECG waveform provides the basis for diagnosis and therapy of numerous conditions associated with disorders of the myocardium in the intensive care unit (ICU) [1]. Continuous ECG monitoring is essential for the adult patient with acute myocardial infarction because dysrhythmias are the most common life-threatening complication [1]. Continuous ECG monitoring of the postoperative general surgical patient is not as useful because the incidence of lifethreatening dysrhythmias is lower [1,2]. However, in adults and children with decreased myocardial function, hypovolemia, or hypoxemia, the ECG may show PVCs, PACs, ST segments changes, altered T-waves, and bradycardia [1]. These abnormalities may be produced by inadequate oxygen delivery to the myocardium and therefore serve to alarm the physician of potentially signi~cant pathophysiologic disturbances. One of the main uses for the ECG is the determination of heart rate or pulse. Most ICUs measure heart rate as a calculated time-averaged signal from the bedside ECG monitor. Heart rate is a very nonspeci~c cardiorespiratory variable. However, tachycardia may signal signi~cant and abrupt cardiovascular compromise such as hypovolemia, hypoperfusion, or diminished myocardial contractility in the critically ill patient [3]. The degree of tachycardia is roughly proportional to the severity of the cardiovascular impairment. Tachycardia is also present in association with infection, anxiety, stress, fever, exercise, pain, and anxiety. Bradycardia may occur with inferior myocardial infarction, arteriosclerotic heart disease, or severe systemic hypoxemia [3]. Heart rate may also be used during invasive hemodynamic monitoring with thermodilution pulmonary arterial catheters to calculate stroke volume and other hemodynamic variables. The ECG may be used directly or as an adjunctive test in the diagnosis of numerous diseases and conditions affecting the critically ill patient. Myocardial ischemia may occur in the critically ill patient with shock, even in the absence of primary coronary artery disease [1]. Myocardial ischemia ultimately results in T-wave inversion. Myocardial injury due to ischemia may cause elevation of the ST segment if the ischemia is transmural or ST segment depression if the ischemia is nontransmural and predominantly subendocardial [1]. Progressive ischemia results in myocardial infarction which is often manifested in the ECG by Qwave deepening and distortion, loss of R-wave voltage, and/or an increased R-wave amplitude [1]. However, ECG abnormalities characterizing myocardial ischemia and infarction are nonspeci~c. Serial ECGs should be followed to speci~cally diagnose and localize myocardial ischemia and infarction [1]. Table 1 lists characteristic changes in the ECG associated with other disease states commonly seen in critically ill patients. ECG changes observed in drug overdose or poisoning, and electrolyte and metabolic disturbances are shown in Tables 2 and 3. Abnormal cardiac rhythms may occur during critical illness at any time and are the most common mechanism of death in adults [1]. Automated arrhythmia detection algorithms are commercially available and in use in many ICUs. These algorithms ~lter the incoming ECG signal, detect and classify the QRS complex, identify ectopic events and ventricular rhythms, and generate alarms. For ventricular ectopic beats, the system has a reported sensitivity of up to 90% with a false positive rate of #1%. However, the algorithms may result in false positive alarms if the Tor P-waves are $50% of the R-wave amplitude, and are not useful for detection of aberrantly conducted beats, atrial ~brillation and _utter, and intermittent bundle branch block. In summary, the clinical utility and cost effectiveness of these and other automated forms of ECG monitoring in the ICU, such as automated ST segment monitoring [5] and signal-averaged electrocardiography (detailed in subsequent sections), have yet to be determined. The ECG waveform itself may be an important source of clinical information that has only recently

Automated accurate emotion recognition system using rhythm-specific deep convolutional neural network technique with multi-channel EEG signals

Electrocardiographic poor R-wave progression: analysis of multiple criteria reveals little usefulness

Prediction of Early Complications in Patients With Acute Myocardial Infarction by Calculation of the ST Score

The electrocardiogram (ECG) is a non-invasive procedure used to diagnose cardiovascular disorders. The 2D ECG system gives important information on the electrophysiology of heart illnesses and possible brain changes, as well as important information about the cardiovascular system’s functional characteristics. The classification proposed system detected four different classes from the ECG cases: myocardial infarction (MI), Arrhythmia, History MI, and Normal case. The suggested system used three steps for classification ECG, include: Image processing, feature extraction, and classification. The first step involves utilizing Image preprocessing methods to prepare and enhance the input images, which includes two steps: converting RGB images to a binary image and applying filters to reduce noise from a binary image. Utilizing local binary pattern algorithm, features are retrieved from the images in the second step (LBP). The collected characteristics are then supplied into a 1D "convolutional neural network" (1D-CNN) model for object detection and classification in the third step. The findings demonstrate that the suggested model of the CNN has a high accuracy rate, with a classification accuracy rate of 97 percent. The suggested method works extremely well in the categorization of the 2D ECG, as demonstrated by the low rate of false alarms and high rate of accuracy.

AimsAcute myocardial infarction (AMI) remains a leading global cause of mortality, where timely diagnosis is critical to enable early intervention. The 12-lead electrocardiogram (ECG) is a critical tool for AMI detection. While deep learning (DL) models show promise for automated ECG analysis, most prior studies rely on small, curated datasets with limited external validation, limiting their clinical applicability.Methods and resultsWe developed a multimodal DL model (Conv-BiLSTM-Attn) integrating convolutional and recurrent neural networks with an attention mechanism. Using a large, multi-centre dataset of 145 656 ECGs from 96 813 patients across three Swedish hospitals. We trained the model to detect AMI using raw 12-lead ECG signals and demographic inputs (age, sex). Model performance was evaluated under two external validation protocols: generalization across hospitals (GAH) and leave-one-hospital-out (LOHO). The model achieved an area under the receiver operating characteristic (AUROC) of 0.848 (95% CI: 0.84–0.86) and an area under the precision-recall (AUPRC) of 0.456, reflecting class imbalance (∼6–10% AMI prevalence) under the GAH protocol. Subgroup AUROCs ranged from 0.79 to 0.92 across age and sex groups. At the Youden-optimized threshold (0.439), the model showed sensitivity of 0.736, specificity of 0.793, negative predictive value of 0.976, and weighted F1-score of 0.837. Under the LOHO protocol, AUROCs ranged from 0.801–0.849. At Youden thresholds (0.34–0.50), sensitivity ranged from 0.671 to 0.776 and specificity from 0.651 to 0.801, confirming generalizability across sites. Conv-BiLSTM-Attn outperformed benchmark models, with Score-CAM highlighting relevant ST–T segments.ConclusionThis DL model can support accurate and generalizable AMI detection from routine ECGs, with the Conv-BiLSTM-Attn architecture outperforming current benchmark approaches.

Heart diseases are among the most common causes of death in the world. Myocardial infarction (MI) requires fast detection with electrocardiogram (ECG) readings. In this paper, we present an end-to-end deep learning system for automated MI and ischemic abnormality detection based on ECG readings with explainability and web deployment. Three models of convolutional neural networks (CNNs) have been trained with one-dimensional inputs from the PTB-XL dataset (21,799 entries, 500 Hz, 10 seconds): 12-Lead CNN with multi-lead digital ECG input, Lead II CNN with single-lead ECG input, and Image CNN with scanned ECG images. 12-Lead CNN outperformed other models with ROC-AUC = 0.8991, sensitivity = 0.8674, specificity = 0.7580, and F1 score = 0.8436. Explainability with Grad-CAM [16] and saliency maps has been incorporated that conforms with the standard ECG pattern interpretation. All of our models have been integrated in an easily accessible web application with Streamlit Cloud to allow WFDB file, CSV, and image input with automatic route selection and report generation.