Electrocardiogram Waveform Research Articles

Multilayered carbon nanotube/adhesive films for human body signal detection sensors

This paper presents the development of multilayered carbon nanotube (CNT)/adhesive (MLCA) films for human body signal detection sensors using a spray-coating method. The chemical composition, adhesion properties, and electrical conductivity of the films were investigated using various adhesives, with the acrylate-based adhesive (ABA) exhibiting superior performance. The surface roughness, thickness, and electrical properties of the films were characterized, and the tunability was demonstrated by adjusting the number of coating layers. Tribological tests were performed to assess the wear resistance and friction behavior of the films. The adhesion stabilities and conformabilities of the films on various substrates were investigated. The films were combined with polydimethylsiloxane (PDMS) and surfactants to create biocompatible and durable sensors. The PDMS-surfactant composite was characterized, and the MLCA film/PDMS-surfactant-based sensor exhibited excellent stability under deformation and biocompatibility. The impedance behavior, temperature, humidity, and strain-sensing capabilities of the sensors were evaluated. The capability of the sensor to detect vital signs was validated by accurately capturing the electrocardiogram (ECG) waveforms. This study provides valuable insights into the design and fabrication of CNT-based conductive films for human body signal-detection sensors, offering a promising approach for the development of flexible and wearable electronic devices.

Study on the Anti-Interference Performance of Substrate-Free PEDOT:PSS ECG Electrodes

Substrate-free electrodes are promising dry electrodes for long-term physiological electrical signal monitoring due to their ultra-thinness, conformal contact, and stable skin–electrode impedance. However, the response of substrate-free electrodes to various disturbances during electrocardiogram (ECG) monitoring and the corresponding optimization needs to be investigated. This paper investigates the specific effects of various influencing factors on skin–electrode impedance and ECG during electrocardiogram (ECG) detection. The research utilizes substrate-free poly(3,4-ethylenedioxythiophene)/poly(styrene-sulfonate) (PEDOT:PSS) electrodes. The investigation employs several methods, including skin–electrode impedance comparison, ECG waveform analysis, spectrum analysis, and signal-to-noise ratio (SNR) evaluation. To avoid the impact of physiological state differences in subjects at different times, relevant data were only compared with the same group of experiments conducted in the same period. The results demonstrate that the substrate-free conformal contact PEDOT:PSS electrode has more stable skin–electrode impedance and could obtain a more stable ECG than partial contact electrodes (the SNR of the partial contact and conformal contact electrodes are 1.2768 ± 4.0299 dB and 7.2637 ± 1.4897 dB, respectively). Furthermore, the ECG signal quality of the substrate-free conformal contact PEDOT:PSS electrode was independent of the electrode area and shape (the SNRs of the large, medium, and small electrodes are 4.0447 ± 0.4616 dB, 3.9115 ± 0.5885 dB, and 4.1556 ± 0.5557 dB, respectively; the SNRs of the circular, square, and triangular electrodes are 9.2649 ± 0.6326 dB, 9.2471 ± 0.6806 dB, and 9.1514 ± 0.6875 dB, respectively), showing high signal acquisition capability that is the same as microneedle electrodes and better than fabric electrodes. The results of clothing friction effects show that skin–electrode impedance stability was important for ECG stability, while the impedance value was not (the SNRs of friction and non-friction electrodes are 2.4128 ± 7.0784 dB and 9.2164 ± 0.6696 dB, respectively). Moreover, the skin–electrode impedance maintains stability even at a high breathing frequency, but the ECG signal fluctuates at a high breathing frequency. This experiment demonstrates that even when the skin–electrode impedance remains stable, the ECG signal can still be susceptible to interference from other factors. This study suggests that substrate-free PEDOT:PSS that could form conformal contact with the skin has higher skin–electrode impedance stability and could measure a high ECG signal even with a small electrode area, demonstrating its potential as dry ECG electrodes, but the interference from other physiological electrical signals may require better circuit design.

Android-Based Wireless Single-Lead Electrocardiogram: Heart Rate Measurement and ECG Signal Visualization

Heart rate (HR) is vital for medical and healthcare purposes. This study presents an Android-based heart rate measurement system utilizing a single-lead electrocardiogram (ECG). Three electrodes placed on the arm in lead I configuration capture the ECG signals. An AD8232 sensor amplifies the signal, which is then digitized by Arduino Nano and transmitted to an Android device via HC-05 Bluetooth. The Android application processes the ECG data using the Pan-Tompkins algorithm with an optimized threshold coefficient to extract HR information. The system displays the ECG waveform and the calculated HR on the user interface. Our evaluation demonstrates high accuracy with an error rate of only 0.042%, sensitivity of 99.84%, and positive predictive value of 97.06%. This research suggests the potential of this system for convenient and reliable HR monitoring using readily available smartphones.

Arrhythmia detection in inter-patient ECG signals using entropy rate features and RR intervals with CNN architecture

The classification of inter-patient ECG data for arrhythmia detection using electrocardiogram (ECG) signals presents a significant challenge. Despite the recent surge in deep learning approaches, there remains a noticeable gap in the performance of inter-patient ECG classification. In this study, we introduce an innovative approach for ECG classification in arrhythmia detection by employing a 1D convolutional neural network (CNN) to leverage both morphological and temporal characteristics of cardiac cycles. Through the utilization of 1D-CNN layers, we automatically capture the morphological attributes of ECG data, allowing us to represent the shape of the ECG waveform around the R peaks. Additionally, we incorporate four RR interval features to provide temporal context, and we explore the potential application of entropy rate as a feature extraction technique for ECG signal classification. Consequently, the classification layers benefit from the combination of both temporal and learned features, leading to the achievement of the final arrhythmia classification. We validate our approach using the MIT-BIH arrhythmia dataset, employing both intra-patient and inter-patient paradigms for model training and testing. The model’s generalization ability is assessed by evaluating it on the INCART dataset. The model attains average accuracy rates of 99.13% and 99.17% for 2-fold and 5-fold cross-validation, respectively, in intra-patient classification with five classes. In inter-patient classification with three and five classes, the model achieves average accuracies of 98.73% and 97.91%, respectively. For the INCART dataset, the model achieves an average accuracy of 98.20% for three classes. The experimental outcomes demonstrate the superiority of the proposed model compared to state-of-the-art models in recognizing arrhythmias. Thus, the proposed model exhibits enhanced generalization and the potential to serve as an effective solution for recognizing arrhythmias in real-world datasets characterized by class imbalances in practical applications.

Non-invasive pulse arrival time is associated with cardiac index in pediatric heart transplant patients with normal ejection fraction

Objective. Cardiac Index (CI) is a key physiologic parameter to ensure end organ perfusion in the pediatric intensive care unit (PICU). Determination of CI requires invasive cardiac measurements and is not routinely done at the PICU bedside. To date, there is no gold standard non-invasive means to determine CI. This study aims to use a novel non-invasive methodology, based on routine continuous physiologic data, called Pulse Arrival Time (PAT) as a surrogate for CI in patients with normal Ejection Fraction (EF). Approach. Electrocardiogram (ECG) and photoplethysmogram (PPG) signals were collected from beside monitors at a sampling frequency of 250 samples per second. Continuous PAT, derived from the ECG and PPG waveforms was averaged per patient. Pearson’s correlation coefficient was calculated between PAT and CI, PAT and heart rate (HR), and PAT and EF. Main Results. Twenty patients underwent right heart cardiac catheterization. The mean age of patients was 11.7 ± 5.4 years old, ranging from 11 months old to 19 years old, the median age was 13.4 years old. HR in this cohort was 93.8 ± 17.0 beats per minute. The average EF was 54.4 ± 9.6%. The average CI was 3.51 ± 0.72 l min−1 m−2, with ranging from 2.6 to 4.77 l min−1 m−2. The average PAT was 0.31 ± 0.12 s. Pearson correlation analysis showed a positive correlation between PAT and CI (0.57, p < 0.01). Pearson correlation between HR and CI, and correlation between EF and CI was 0.22 (p = 0.35) and 0.03 (p = 0.23) respectively. The correlation between PAT, when indexed by HR (i.e. PAT × HR), and CI minimally improved to 0.58 (p < 0.01). Significance. This pilot study demonstrates that PAT may serve as a valuable surrogate marker for CI at the bedside, as a non-invasive and continuous modality in the PICU. The use of PAT in clinical practice remains to be thoroughly investigated.

Multimodal ChatGPT-4V for Electrocardiogram Interpretation: Promise and Limitations.

This study evaluated the capabilities of the newly released ChatGPT-4V, a large language model with visual recognition abilities, in interpreting electrocardiogram waveforms and answering related multiple-choice questions for assisting with cardiovascular care.

Data imbalance in cardiac health diagnostics using CECG-GAN

Heart disease is the world’s leading cause of death. Diagnostic models based on electrocardiograms (ECGs) are often limited by the scarcity of high-quality data and issues of data imbalance. To address these challenges, we propose a conditional generative adversarial network (CECG-GAN). This strategy enables the generation of samples that closely approximate the distribution of ECG data. Additionally, CECG-GAN addresses waveform jitter, slow processing speeds, and dataset imbalance issues through the integration of a transformer architecture. We evaluated this approach using two datasets: MIT-BIH and CSPC2020. The experimental results demonstrate that CECG-GAN achieves outstanding performance metrics. Notably, the percentage root mean square difference (PRD) reached 55.048, indicating a high degree of similarity between generated and actual ECG waveforms. Additionally, the Fréchet distance (FD) was approximately 1.139, the root mean square error (RMSE) registered at 0.232, and the mean absolute error (MAE) was recorded at 0.166.

Fetal electrocardiogram signal extraction based on multi-scale residual shrinkage U-Net

In the extraction of fetal electrocardiogram (ECG) signal, due to the unicity of the scale of the U-Net same-level convolution encoder, the size and shape difference of the ECG characteristic wave between mother and fetus are ignored, and the time information of ECG signals is not used in the threshold learning process of the encoder's residual shrinkage module. In this paper, a method of extracting fetal ECG signal based on multi-scale residual shrinkage U-Net model is proposed. First, the Inception and time domain attention were introduced into the residual shrinkage module to enhance the multi-scale feature extraction ability of the same level convolution encoder and the utilization of the time domain information of fetal ECG signal. In order to maintain more local details of ECG waveform, the maximum pooling in U-Net was replaced by Softpool. Finally, the decoder composed of the residual module and up-sampling gradually generated fetal ECG signals. In this paper, clinical ECG signals were used for experiments. The final results showed that compared with other fetal ECG extraction algorithms, the method proposed in this paper could extract clearer fetal ECG signals. The sensitivity, positive predictive value, and F1 scores in the 2013 competition data set reached 93.33%, 99.36%, and 96.09%, respectively, indicating that this method can effectively extract fetal ECG signals and has certain application values for perinatal fetal health monitoring.

Machine learning of ECG waveforms and cardiac magnetic resonance for response and survival after cardiac resynchronization therapy

Cardiac resynchronization therapy (CRT) can lead to marked symptom reduction and improved survival in selected patients with heart failure with reduced ejection fraction (HFrEF); however, many candidates for CRT based on clinical guidelines do not have a favorable response. A better way to identify patients expected to benefit from CRT that applies machine learning to accessible and cost-effective diagnostic tools such as the 12-lead electrocardiogram (ECG) could have a major impact on clinical care in HFrEF by helping providers personalize treatment strategies and avoid delays in initiation of other potentially beneficial treatments. This study addresses this need by demonstrating that a novel approach to ECG waveform analysis using functional principal component decomposition (FPCD) performs better than measures that require manual ECG analysis with the human eye and also at least as well as a previously validated but more expensive approach based on cardiac magnetic resonance (CMR). Analyses are based on five-fold cross validation of areas under the curve (AUCs) for CRT response and survival time after the CRT implant using Cox proportional hazards regression with stratification of groups using a Gaussian mixture model approach. Furthermore, FPCD and CMR predictors are shown to be independent, which demonstrates that the FPCD electrical findings and the CMR mechanical findings together provide a synergistic model for response and survival after CRT. In summary, this study provides a highly effective approach to prognostication after CRT in HFrEF using an accessible and inexpensive diagnostic test with a major expected impact on personalization of therapies.

Restoration of magnetohydrodynamic-corrupted 12-lead electrocardiogram to enhance cardiac monitoring during magnetic resonance imaging

The Magnetohydrodynamic (MHD) effect on the bloodstream, induced by the static magnetic field of Magnetic Resonance Imaging (MRI) devices, distorts Electrocardiogram (ECG) components, and poses challenges to cardiac gating and monitoring during MRI examinations. Restoring ECG components from MHD-induced artifacts has always been a challenge, primarily due to the influence of numerous device-induced and physiological variables and the absence of paired ground truth clean ECG for validation. In this research, at first, we employ unpaired restoration of MHD-corrupted 12-lead ECG signals using one dimensional (1D) Cycle Generative Adversarial Networks (CycleGANs) for channel and patient orientation-specific restoration. Subsequently, a 1D-segmentation-based comprehensive restoration scheme is proposed based on the phase 1 outputs to simultaneously restore 12-channel ECG signals, irrespective of internal and external factors affecting ECG. By employing the proposed Magnetohydrodynamic Generative Adversarial Network (MHDGAN) and MHD-to-ECG Network (MHD2ECGNet) for unpaired and comprehensive restoration, we achieved an overall spectral correlation improvement of 92.56% and 77.64%, respectively. In R-peak detection, MHDGAN demonstrated an overall precision, recall, and F1-score of 98.14%, 94.12%, and 96.01%, respectively. The extracted heart rate (HR) and heart rate variability (HRV) from the restored waveforms closely match the ground truth, as confirmed through quantitative and qualitative assessments. The proposed framework has proven effective in restoring MHD-corrupted ECG waveforms, even in the absence of paired ground truth labels and amidst compounding distortions from multiple sources.

Enhanced Cardiovascular Assessment in Rats Using a Wireless Data Logger

Study objectives and hypothesis: Cardiovascular assessment in animals can be complicated by signal loss of real-time telemetry. We evaluated the benefits of using a wireless data logger to capture 18 hours of blood pressure (BP) and electrocardiogram (ECG) data. This approach eliminates data loss, simplifies analysis and improves results. Methods: One male Sprague Dawley rat was implanted with a PBTA-L implant (Stellar, TSE Systems) using established procedures to collect BP and ECG. All procedures were approved by the AmplifyBio IACUC. Memory and real-time (RT) BP and ECG were recorded at 500Hz for 18+ hours in the acquisition software (Notocord-hem, Instem) on separate days. ECG RR intervals and BP intervals were compared for memory versus RT recordings. Time-based (RMSSD, SDNN, SDANN) and frequency-based (LF, HF) HRV assessment was performed in 5-minute increments. HRV from ECG vs BP were compared for the memory recording. Data and Results: High quality BP and ECG waveforms were recorded to memory and retrieved into the software with no manual intervention or loss of data. The RT recording included 0.48% signal loss. Interval outliers were removed for HRV assessment using thresholds to limit valid RR/beat intervals. This worked well for BP but was less effective for ECG as there were more erroneous intervals due to movement artifact. Time- and frequency-based HRV parameters were compared for memory vs RT recordings for BP and ECG intervals. HRV from BP showed lower variability for RT vs memory (LF↓24.3%, HR↓6.3%, RMSSD↓0.1%, SDNN↓15.4%, SDANN↓6.1%), while HRV from ECG generally produced higher variability for RT vs memory (LF↑111.8%, HF↑175.1%, RMSSD↑93.9%, SDNN↑16.7%, SDANN↓9.5%). These findings can be explained as follows: BP intervals produced less variability with the RT recording because the long intervals caused by signal loss are distinct and consistently removed by threshold criteria. The lost intervals result in lower HRV. ECG intervals produced more variability with the RT recording because the recordings were 1 week apart, and the later RT recording had more movement artifact on the ECG resulting in inaccurate QRS detection. The erroneous intervals were not as effectively removed by thresholds, resulting in higher HRV. We also compared HRV from the BP and ECG signals in the memory recording. The LF and HF powers were 5.4 and 4.5 times as high respectively for the RR (ECG) vs beat (BP) intervals. The time-based HRV parameters for RMSSD and SDNN were higher by 84% and 14%, while SDANN was very similar with ECG results 2% less than BP results. Conclusions: HRV analysis highlights the benefits of using a wireless data logger to record cardiovascular data. Our assessment demonstrated improvements in the effciency and accuracy of HRV analysis. As a secondary outcome, we demonstrated potential benefits of performing HRV analysis on BP vs ECG signals. None. This is the full abstract presented at the American Physiology Summit 2024 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.

Comparison of electrocardiogram (ECG) waveforms and centralized ECG measurements between a wearable mobile ECG device and a standard 12-lead ECG

Comparison of electrocardiogram (ECG) waveforms and centralized ECG measurements between a wearable mobile ECG device and a standard 12-lead ECG

ECG-surv: A deep learning-based model to predict time to 1-year mortality from 12-lead electrocardiogram

Electrocardiogram (ECG) abnormalities have demonstrated potential as prognostic indicators of patient survival. However, the traditional statistical approach is constrained by structured data input, limiting its ability to fully leverage the predictive value of ECG data in prognostic modeling. This study aims to introduce and evaluate a deep-learning model to simultaneously handle censored data and unstructured ECG data for survival analysis. We herein introduce a novel deep neural network called ECG-surv, which includes a feature extraction neural network and a time-to-event analysis neural network. The proposed model is specifically designed to predict the time to 1-year mortality by extracting and analyzing unique features from 12-lead ECG data. ECG-surv was evaluated using both an independent test set and an external set, which were collected using different ECG devices. The performance of ECG-surv surpassed that of the Cox proportional model, which included demographics and ECG waveform parameters, in predicting 1-year all-cause mortality, with a significantly higher concordance index (C-index) in ECG-surv than in the Cox model using both the independent test set (0.860 [95% CI: 0.859- 0.861] vs. 0.796 [95% CI: 0.791- 0.800]) and the external test set (0.813 [95% CI: 0.807- 0.814] vs. 0.764 [95% CI: 0.755- 0.770]). ECG-surv also demonstrated exceptional predictive ability for cardiovascular death (C-index of 0.891 [95% CI: 0.890- 0.893]), outperforming the Framingham risk Cox model (C-index of 0.734 [95% CI: 0.715-0.752]). ECG-surv effectively utilized unstructured ECG data in a survival analysis. It outperformed traditional statistical approaches in predicting 1-year all-cause mortality and cardiovascular death, which makes it a valuable tool for predicting patient survival.

Simple models vs. deep learning in detecting low ejection fraction from the electrocardiogram.

Deep learning methods have recently gained success in detecting left ventricular systolic dysfunction (LVSD) from electrocardiogram (ECG) waveforms. Despite their high level of accuracy, they are difficult to interpret and deploy broadly in the clinical setting. In this study, we set out to determine whether simpler models based on standard ECG measurements could detect LVSD with similar accuracy to that of deep learning models. Using an observational data set of 40 994 matched 12-lead ECGs and transthoracic echocardiograms, we trained a range of models with increasing complexity to detect LVSD based on ECG waveforms and derived measurements. The training data were acquired from the Stanford University Medical Center. External validation data were acquired from the Columbia Medical Center and the UK Biobank. The Stanford data set consisted of 40 994 matched ECGs and echocardiograms, of which 9.72% had LVSD. A random forest model using 555 discrete, automated measurements achieved an area under the receiver operator characteristic curve (AUC) of 0.92 (0.91-0.93), similar to a deep learning waveform model with an AUC of 0.94 (0.93-0.94). A logistic regression model based on five measurements achieved high performance [AUC of 0.86 (0.85-0.87)], close to a deep learning model and better than N-terminal prohormone brain natriuretic peptide (NT-proBNP). Finally, we found that simpler models were more portable across sites, with experiments at two independent, external sites. Our study demonstrates the value of simple electrocardiographic models that perform nearly as well as deep learning models, while being much easier to implement and interpret.

Study of Attenuation Correction Using a Cardiac Dynamic Phantom: Synchronized Time-Phase-Gated Attenuation Correction Method.

In cardiac nuclear medicine examinations, absorption in the body is the main factor in the degradation of the image quality. The Chang and external source methods were used to correct for absorption in the body. However, fundamental studies on attenuation correction for electrocardiogram (ECG)-synchronized CT imaging have not been performed. Therefore, we developed and improved an ECG-synchronized cardiac dynamic phantom and investigated the synchronized time-phase-gated attenuation correction (STPGAC) method using ECG-synchronized SPECT and CT images of the same time phase. Methods: As a basic study, SPECT was performed using synchronized time-phase-gated (STPG) SPECT and non-phase-gated (NPG) SPECT. The attenuation-corrected images were, first, CT images with the same time phase as the ECG waveform of the gated SPECT acquisition (with CT images with the ECG waveform of the CT acquisition as the reference); second, CT images with asynchronous ECG; third, CT images of the 75% region; and fourth, CT images of the 40% region. Results: In the analysis of cardiac function in the phantom experiment, left ventricle ejection fraction (heart rate, 11.5%-13.4%; myocardial wall, 49.8%-55.7%) in the CT images was compared with that in the STPGAC method (heart rate, 11.5%-13.3%; myocardial wall, 49.6%-55.5%), which was closer in value to that of the STPGAC method. In the phantom polar map segment analyses, none of the images showed variability (F (10,10) < 0.5, P = 0.05). All images were correlated (r = 0.824-1.00). Conclusion: In this study, we investigated the STPGAC method using a SPECT/CT system. The STPGAC method showed similar values of cardiac function analysis to the CT images, suggesting that the STPGAC method accurately reconstructed the distribution of blood flow in the myocardial region. However, the target area for attenuation correction of the heart region was smaller than that of the whole body, and changing the gated SPECT conditions and attenuation-corrected images did not affect myocardial blood flow analysis.

On‐Skin Paintable Water‐Resistant Biohydrogel for Wearable Bioelectronics

AbstractTo achieve accurate monitoring of bioelectrical signals, it is essential to use customizable bioelectrodes that can self‐adapt to the skin's surface topography. Ion‐conducting hydrogel has received significant attention in this field due to its softness, adhesion, and skin‐like mechanical properties. However, these bioelectrodes currently suffer from degradation of adhesion, electrical conductivity, and skin‐compliance when exposed to aqueous environments. This significantly limits the application of bioelectrodes. Herein, a customizable biohydrogel that can be applied on skin or fabric by solvent volatilization for liquid ink‐gel film conversion is reported. The biohydrogel's distinct characteristic of transitioning between a liquid and hydrogel phase establishes superb conformal contact and dynamic compliance with the epidermis. This effectively eliminates motion artifacts and results in lower contact impedance and noise in both static and dynamic states when compared to existing bioelectrodes. The biohydrogel is applied to the cotton fabric to create electrocardiogram (ECG) monitoring garments. These garments enable the acquisition of ECG signals with high accuracy in aqueous environment for over 72 h. Besides, the biohydrogel‐based garments outperform the commercial gel electrodes by 83.5% in signal‐to‐noise ratio. Additionally, the cotton/biohydrogel electrode facilitates multi‐channel, high‐fidelity recording of ECG signals, enabling high‐performance capture and classification of ECG waveforms across multiple channels.

Relationship between motion artifacts and fiber length of a flocked electrode in electrocardiogram measurement

In this paper, we report on the relationship between the fiber length of flocked electrodes and the motion artifacts (MAs) that occur in electrocardiogram (ECG) measurements. During ECG measurements, fluctuations in the ECG baseline are observed when slippage occurs at the skin-to-flocked electrode interface. We evaluated the friction behavior and MAs characteristics of four types of flocked electrode with various fiber lengths (0.3 mm, 0.5 mm, 0.7 mm, and 1.2 mm). As a result, it was confirmed that the longer fiber lengths for the flocked electrodes provided more stable contact with the skin phantom. The correlation coefficients between the ECG waveform without MA and the ECG waveform using each fiber length (0.3 mm, 0.5 mm, 0.7 mm, 1.2 mm) were 0.58, 0.58, 0.81, and 0.91. Since the correlation coefficient close to 1 means less MA, it was shown that the longer fiber length, the higher the MA reduction effect. We also conducted an ECG measurement experiment using an ECG measurement chest band. When the flocked electrode with the 1.2-mm-long fibers was used, the ECG signal was measured successfully with only small MAs from body movements caused by respiration.

Use of the energy waveform electrocardiogram to detect subclinical left ventricular dysfunction in patients with type 2 diabetes mellitus

BackgroundRecent guidelines propose N-terminal pro-B-type natriuretic peptide (NT-proBNP) for recognition of asymptomatic left ventricular (LV) dysfunction (Stage B Heart Failure, SBHF) in type 2 diabetes mellitus (T2DM). Wavelet Transform based signal-processing transforms electrocardiogram (ECG) waveforms into an energy distribution waveform (ew)ECG, providing frequency and energy features that machine learning can use as additional inputs to improve the identification of SBHF. Accordingly, we sought whether machine learning model based on ewECG features was superior to NT-proBNP, as well as a conventional screening tool—the Atherosclerosis Risk in Communities (ARIC) HF risk score, in SBHF screening among patients with T2DM.MethodsParticipants in two clinical trials of SBHF (defined as diastolic dysfunction [DD], reduced global longitudinal strain [GLS ≤ 18%] or LV hypertrophy [LVH]) in T2DM underwent 12-lead ECG with additional ewECG feature and echocardiography. Supervised machine learning was adopted to identify the optimal combination of ewECG extracted features for SBHF screening in 178 participants in one trial and tested in 97 participants in the other trial. The accuracy of the ewECG model in SBHF screening was compared with NT-proBNP and ARIC HF.ResultsSBHF was identified in 128 (72%) participants in the training dataset (median 72 years, 41% female) and 64 (66%) in the validation dataset (median 70 years, 43% female). Fifteen ewECG features showed an area under the curve (AUC) of 0.81 (95% CI 0.787–0.794) in identifying SBHF, significantly better than both NT-proBNP (AUC 0.56, 95% CI 0.44–0.68, p < 0.001) and ARIC HF (AUC 0.67, 95%CI 0.56–0.79, p = 0.002). ewECG features were also led to robust models screening for DD (AUC 0.74, 95% CI 0.73–0.74), reduced GLS (AUC 0.76, 95% CI 0.73–0.74) and LVH (AUC 0.90, 95% CI 0.88–0.89).ConclusionsMachine learning based modelling using additional ewECG extracted features are superior to NT-proBNP and ARIC HF in SBHF screening among patients with T2DM, providing an alternative HF screening strategy for asymptomatic patients and potentially act as a guidance tool to determine those who required echocardiogram to confirm diagnosis.Trial registration LEAVE-DM, ACTRN 12619001393145 and Vic-ELF, ACTRN 12617000116325

A comprehensive review on efficient artificial intelligence models for classification of abnormal cardiac rhythms using electrocardiograms

Deep learning has made many advances in data classification using electrocardiogram (ECG) waveforms. Over the past decade, data science research has focused on developing artificial intelligence (AI) based models that can analyze ECG waveforms to identify and classify abnormal cardiac rhythms accurately. However, the primary drawback of the current AI models is that most of these models are heavy, computationally intensive, and inefficient in terms of cost for real-time implementation. In this review, we first discuss the current state-of-the-art AI models utilized for ECG-based cardiac rhythm classification. Next, we present some of the upcoming modeling methodologies which have the potential to perform real-time implementation of AI-based heart rhythm diagnosis. These models hold significant promise in being lightweight and computationally efficient without compromising the accuracy. Contemporary models predominantly utilize 12-lead ECG for cardiac rhythm classification and cardiovascular status prediction, increasing the computational burden and making real-time implementation challenging. We also summarize research studies evaluating the potential of efficient data setups to reduce the number of ECG leads without affecting classification accuracy. Lastly, we present future perspectives on AI's utility in precision medicine by providing opportunities for accurate prediction and diagnostics of cardiovascular status in patients.

One-shot screening: Utilization of a two-dimensional convolutional neural network for automatic detection of left ventricular hypertrophy using electrocardiograms

Background and ObjectiveLeft ventricular hypertrophy (LVH) can impair ejection function and elevate the risk of heart failure. Therefore, early detection through screening is crucial. This study aimed to propose a novel method to enhance LVH detection using 12-lead electrocardiogram (ECG) waveforms with a two-dimensional (2D) convolutional neural network (CNN). MethodsUtilizing 42,127 pairs of ECG-transthoracic echocardiogram data, we pre-processed raw data into single-shot images derived from each ECG lead and conducted lead selection to optimize LVH diagnosis. Our proposed one-shot screening method, implemented during pre-processing, enables the superimposition of waveform source data of any length onto a single-frame image, thereby addressing the limitations of the one-dimensional (1D) approach. We developed a deep learning model with a 2D-CNN structure and machine learning models for LVH detection. To assess our method, we also compared our results with conventional ECG criteria and those of a prior study that used a 1D-CNN approach, utilizing the same dataset from the University of Tokyo Hospital for LVH diagnosis. ResultsFor LVH detection, the average area under the receiver operating characteristic curve (AUROC) was 0.916 for the 2D-CNN model, which was significantly higher than that obtained using logistic regression and random forest methods, as well as the two conventional ECG criteria (AUROC of 0.766, 0.790, 0.599, and 0.622, respectively). Incorporating additional metadata, such as ECG measurement data, further improved the average AUROC to 0.921. The model's performance remained stable across two different annotation criteria and demonstrated significant superiority over the performance of the 1D-CNN model used in a previous study (AUROC of 0.807). ConclusionsThis study introduces a robust and computationally efficient method that outperforms 1D-CNN models utilized in previous studies for LVH detection. Our method can transform waveforms of any length into fixed-size images and leverage the selected lead of the ECG, ensuring adaptability in environments with limited computational resources. The proposed method holds promise for integration into clinical practice as a tool for early diagnosis, potentially enhancing patient outcomes by facilitating earlier treatment and management.