Automatic Detection of Atrial Fibrillation from Single-Lead ECG Using Deep Learning of the Cardiac Cycle

Blood Pressure Estimation Based on PPG and ECG Signals Using Knowledge Distillation

OBJECTIVE

Easy access bio-signals are useful for alleviating the shortcomings and difficulties associated with cuff-based and invasive blood pressure (BP) measurement techniques. This study proposes a deep learning model, trained using knowledge distillation, based on photoplethysmographic (PPG) and electrocardiogram (ECG) signals to estimate systolic and diastolic blood pressures.

METHODS

The estimation model comprises convolutional layers followed by one bidirectional recurrent layer and attention layers. The training approach involves knowledge distillation, where a smaller model (student model) is trained by leveraging information from a larger model (teacher model).

RESULTS

The proposed multistage model was evaluated on 1205 subjects from Medical Information Mart for Intensive Care (MIMIC) III database using the Association for the Advancement of Medical Instrumentation (AAMI) and the standards of the British Hypertension Society (BHS). The results revealed that our model performance achieved grade A in estimating both systolic blood pressure (SBP) and diastolic blood pressure (DBP) and met the requirements of the AAMI standard. After training with knowledge distillation (KD), the model achieved a mean absolute error and standard deviation of 2.94 ± 5.61 mmHg for SBP and 2.02 ± 3.60 mmHg for DBP.

CONCLUSION

Our results demonstrate the benefits of the knowledge distillation training method in reducing the number of parameters and improving the predictive accuracy of the blood pressure regression model.

Efficient and Automatic Breast Cancer Early Diagnosis System Based on the Hierarchical Extreme Learning Machine

Breast cancer is the leading type of cancer in women, causing nearly 600,000 deaths every year, globally. Although the tumors can be localized within the breast, they can spread to other body parts, causing more harm. Therefore, early diagnosis can help reduce the risks of this cancer. However, a breast cancer diagnosis is complicated, requiring biopsy by various methods, such as MRI, ultrasound, BI-RADS, or even needle aspiration and cytology with the suggestions of specialists. On certain occasions, such as body examinations of a large number of people, it is also a large workload to check the images. Therefore, in this work, we present an efficient and automatic diagnosis system based on the hierarchical extreme learning machine (H-ELM) for breast cancer ultrasound results with high efficiency and make a primary diagnosis of the images. To make it compatible to use, this system consists of PNG images and general medical software within the H-ELM framework, which is easily trained and applied. Furthermore, this system only requires ultrasound images on a small scale, of 28×28 pixels, reducing the resources and fulfilling the application with low-resolution images. The experimental results show that the system can achieve 86.13% in the classification of breast cancer based on ultrasound images from the public breast ultrasound images (BUSI) dataset, without other relative information and supervision, which is higher than the conventional deep learning methods on the same dataset. Moreover, the training time is highly reduced, to only 5.31 s, and consumes few resources. The experimental results indicate that this system could be helpful for precise and efficient early diagnosis of breast cancers with primary examination results.

Diagnosis of atrial fibrillation based on unsupervised domain adaptation

In recent years, the proportion of the elderly in the society is continuously increasing. Cardiovascular disease is a big problem that puzzles the health of the elderly. Among them, atrial fibrillation is one of the most common arrhythmia diseases in recent years, which poses a great threat to human life safety. At the same time, deep learning has become a powerful tool for medical and healthcare applications due to its high accuracy and fast detection speed. The diagnosis of atrial fibrillation is based on electrocardiogram, ECG) timing signals. At present, the scale of the open ECG data set is limited, and a large amount of labeled ECG data is needed to build a high-precision diagnostic model. In this study, a two-channel network model and a feature queue technique are proposed. A high-quality classification diagnosis model of atrial fibrillation is obtained by unsupervised domain adaptive technique, which uses a small amount of labeled data and a large amount of unlabeled data for training. The research content of this paper includes the following aspects: 1) Build a dual-channel network model, which can analyze ECG signals from different feature dimensions. At the same time, the dual-channel output also improves the reliability of the model's pseudo-label in the adaptive training stage and the accuracy of the output in the testing stage. 2) Innovative feature queue technology including global centroid is proposed to participate in the process of domain discrepancy metric calculation, which can use a small amount of labeled data and a large amount of unlabeled data to achieve a more stable and rapid update of the network. 3) Improved and innovated the domain discrepancy metric function, and introduced an evaluation formula for the credibility of false labels to improve the learning efficiency of unlabeled data. Finally, the experimental results show that the proposed two-channel network model and the feature queue technique with global centroid can achieve a high generalization and high precision depth network model by training with a small amount of labeled data and a large amount of unlabeled data. 4) The proposed model achieved a precision of 95.12%, a recall of 95.36%, an accuracy of 98.05%, and an F1 score of 95.23% in the MIT-BIH Arrhythmia Database. In the MIT-BIH Atrial Fibrillation Database, the model achieved a precision of 98.9%, a recall of 99.03%, an accuracy of 99.13%, and an F1 score of 99.08%.

Prediction of Sleep Apnea Events Using a CNN-Transformer Network and Contactless Breathing Vibration Signals

It is estimated that globally 425 million subjects have moderate to severe obstructive sleep apnea (OSA). The accurate prediction of sleep apnea events can offer insight into the development of treatment therapies. However, research related to this prediction is currently limited. We developed a covert framework for the prediction of sleep apnea events based on low-frequency breathing-induced vibrations obtained from piezoelectric sensors. A CNN-transformer network was utilized to efficiently extract local and global features from respiratory vibration signals for accurate prediction. Our study involved overnight recordings of 105 subjects. In five-fold cross-validation, we achieved an accuracy of 85.9% and an F1 score of 85.8%, which are 3.5% and 5.3% higher than the best-performed classical model, respectively. Additionally, in leave-one-out cross-validation, 2.3% and 3.8% improvements are observed, respectively. Our proposed CNN-transformer model is effective in the prediction of sleep apnea events. Our framework can thus provide a new perspective for improving OSA treatment modes and clinical management.

An Arrhythmia Classification Model Based on Vision Transformer with Deformable Attention

The electrocardiogram (ECG) is a highly effective non-invasive tool for monitoring heart activity and diagnosing cardiovascular diseases (CVDs). Automatic detection of arrhythmia based on ECG plays a critical role in the early prevention and diagnosis of CVDs. In recent years, numerous studies have focused on using deep learning methods to address arrhythmia classification problems. However, the transformer-based neural network in current research still has a limited performance in detecting arrhythmias for the multi-lead ECG. In this study, we propose an end-to-end multi-label arrhythmia classification model for the 12-lead ECG with varied-length recordings. Our model, called CNN-DVIT, is based on a combination of convolutional neural networks (CNNs) with depthwise separable convolution, and a vision transformer structure with deformable attention. Specifically, we introduce the spatial pyramid pooling layer to accept varied-length ECG signals. Experimental results show that our model achieved an F1 score of 82.9% in CPSC-2018. Notably, our CNN-DVIT outperforms the latest transformer-based ECG classification algorithms. Furthermore, ablation experiments reveal that the deformable multi-head attention and depthwise separable convolution are both efficient in extracting features from multi-lead ECG signals for diagnosis. The CNN-DVIT achieved good performance for the automatic arrhythmia detection of ECG signals. This indicates that our research can assist doctors in clinical ECG analysis, providing important support for the diagnosis of arrhythmia and contributing to the development of computer-aided diagnosis technology.

Multifunctional Biosensing Platform Based on Nickel-Modified Laser-Induced Graphene

Nickel plating electrolytes prepared by using a simple salt solution can achieve nickel plating on laser-induced graphene (LIG) electrodes, which greatly enhances the electrical conductivity, electrochemical properties, wear resistance, and corrosion resistance of LIG. This makes the LIG-Ni electrodes well suited for electrophysiological, strain, and electrochemical sensing applications. The investigation of the mechanical properties of the LIG-Ni sensor and the monitoring of pulse, respiration, and swallowing confirmed that the sensor can sense insignificant deformations to relatively large conformal strains of skin. Modulation of the nickel-plating process of LIG-Ni, followed by chemical modification, may allow for the introduction of glucose redox catalyst Ni₂Fe(CN)₆ with interestingly strong catalytic effects, which gives LIG-Ni impressive glucose-sensing properties. Additionally, the chemical modification of LIG-Ni for pH and Na⁺ monitoring also confirmed its strong electrochemical monitoring potential, which demonstrates application prospects in the development of multiple electrochemical sensors for sweat parameters. A more uniform LIG-Ni multi-physiological sensor preparation process provides a prerequisite for the construction of an integrated multi-physiological sensor system. The sensor was validated to have continuous monitoring performance, and its preparation process is expected to form a system for non-invasive physiological parameter signal monitoring, thus contributing to motion monitoring, disease prevention, and disease diagnosis.

Atrioventricular Synchronization for Detection of Atrial Fibrillation and Flutter in One to Twelve ECG Leads Using a Dense Neural Network Classifier

This study investigates the use of atrioventricular (AV) synchronization as an important diagnostic criterion for atrial fibrillation and flutter (AF) using one to twelve ECG leads. Heart rate, lead-specific AV conduction time, and P-/f-wave amplitude were evaluated by three representative ECG metrics (mean value, standard deviation), namely RR-interval (RRi-mean, RRi-std), PQ-interval (PQi-mean, PQI-std), and PQ-amplitude (PQa-mean, PQa-std), in 71,545 standard 12-lead ECG records from the six largest PhysioNet CinC Challenge 2021 databases. Two rhythm classes were considered (AF, non-AF), randomly assigning records into training (70%), validation (20%), and test (10%) datasets. In a grid search of 19, 55, and 83 dense neural network (DenseNet) architectures and five independent training runs, we optimized models for one-lead, six-lead (chest or limb), and twelve-lead input features. Lead-set performance and SHapley Additive exPlanations (SHAP) input feature importance were evaluated on the test set. Optimal DenseNet architectures with the number of neurons in sequential [1st, 2nd, 3rd] hidden layers were assessed for sensitivity and specificity: DenseNet [16,16,0] with primary leads (I or II) had 87.9-88.3 and 90.5-91.5%; DenseNet [32,32,32] with six limb leads had 90.7 and 94.2%; DenseNet [32,32,4] with six chest leads had 92.1 and 93.2%; and DenseNet [128,8,8] with all 12 leads had 91.8 and 95.8%, indicating sensitivity and specificity values, respectively. Mean SHAP values on the entire test set highlighted the importance of RRi-mean (100%), RR-std (84%), and atrial synchronization (40-60%) for the PQa-mean (aVR, I), PQi-std (V2, aVF, II), and PQi-mean (aVL, aVR). Our focus on finding the strongest AV synchronization predictors of AF in 12-lead ECGs would lead to a comprehensive understanding of the decision-making process in advanced neural network classifiers. DenseNet self-learned to rely on a few ECG behavioral characteristics: first, characteristics usually associated with AF conduction such as rapid heart rate, enhanced heart rate variability, and large PQ-interval deviation in V2 and inferior leads (aVF, II); second, characteristics related to a typical P-wave pattern in sinus rhythm, which is best distinguished from AF by the earliest negative P-peak deflection of the right atrium in the lead (aVR) and late positive left atrial deflection in lateral leads (I, aVL). Our results on lead-selection and feature-selection practices for AF detection should be considered for one- to twelve-lead ECG signal processing settings, particularly those measuring heart rate, AV conduction times, and P-/f-wave amplitudes. Performances are limited to the AF diagnostic potential of these three metrics. SHAP value importance can be used in combination with a human expert's ECG interpretation to change the focus from a broad observation of 12-lead ECG morphology to focusing on the few AV synchronization findings strongly predictive of AF or non-AF arrhythmias. Our results are representative of AV synchronization findings across a broad taxonomy of cardiac arrhythmias in large 12-lead ECG databases.