Non-invasive Coronary Artery Disease Screening Based on Electrocardiogram Characteristics and Clinical Risk Factors.

Coronary artery disease (CAD), an acute and life-threatening cardiovascular disease, is a leading cause of mortality and morbidity worldwide. Coronary angiography, the principal diagnostic tool for CAD, is invasive, expensive, and requires a lot of skilled effort. The current study aims to develop an automated and non-invasive CAD detection model and improve its performance as closely as possible to clinically acceptable diagnostic sensitivity. Electrocardiogram (ECG) characteristics are observed to be altered due to CAD and can be studied to develop a screening tool for its detection. The subject's clinical information can help broadly identify the high-cardiac-risk population and serve as a primary step in diagnosing CAD. This paper presents an approach to automatically detect CAD based on clinical data, morphological ECG features, and heart rate variability (HRV) features extracted from short-duration Lead-II ECG recordings. A few popular machine-learning classifiers, including support vector machine (SVM), random forest (RF), K-nearest neighbours (KNN), Gaussian Naïve Bayes (GNB), and multi-layer perceptron (MLP), are trained on the extracted feature space, and their performance is evaluated. Classifiers built by integrating clinical data and features extracted from ECG recordings demonstrated better performance than those built on each feature set separately, and the RF classifier outperforms other considered machine learners and reports an average testing accuracy of 94% and a G-mean score of 92% with a 5-fold cross-validation training accuracy of 95(± 0.04)%.Clinical relevance- The proposed method uses a brief, single-lead ECG recording and performs similarly to current clinical practices in an explainable manner. This makes it suitable for deployment via wearable technology (like smart watch gadgets) and telemonitoring, which may facilitate an earlier and more widespread CAD diagnosis.

Automated Myocardial Infarction Screening Using Morphology-Based Electrocardiogram Biomarkers.

Ischemic heart disease (IHD), a critical and dreadful cardiovascular disease, is a leading cause of death globally. The steady progress of IHD leads to an irreversible condition called myocardial infarction (MI). The detection of MI can be done by observing the altered electrocardiogram (ECG) characteristics. Often, automated ECG analysis is preferred in place of visual inspection to reduce time and ensure reliable detection even when the recording quality is not very good. This paper presents an automated approach to classify recent MI, past MI, and normal sinus rhythm (NSR) classes based on the morphological features of the ECG. In clinical practice, a standard 12-lead ECG setup is typically employed to identify MI. However, acquiring a 12-lead ECG is not always convenient. Hence, in this study, we have explored the possibility of using a minimal number of ECG leads by deriving the augmented limb leads using leads I and II. A well-known and widely used ensemble machine learning tool, the random forest (RF) classifier is trained using features extracted from the derived augmented limb leads and their combinations. An RF classifier built using features extracted from all limb leads has outperformed classifiers built on combinations of them with five-fold cross-validation training accuracy of 97.9 (±0.008) % and testing accuracy of 98 %.Clinical relevance- As high sensitivity is reported in identifying recent MI and past MI classes, the proposed approach is suitable for preventative healthcare applications since it is less likely that subjects with recent or past MI will be misclassified. Due to its low computational complexity, better interpretability, and comparable performance to the state-of-the-art results, the proposed approach can be employed in clinical and cardiac health screening applications. It also has the potential to be employed in remote monitoring with mobile and wearable devices because it is built on features extracted from only lead I and II ECG recordings.

Detection of Twin Pregnancies using Fetal Phonocardiogram.

Fetal phonocardiogram (fPCG), or the electronic recording of fetal heart sounds, is a safe and easily available signal that can be used to monitor fetal wellbeing. In the proposed work an attempt is made to identify twin pregnancies using fPCG data recorded from the fetus with 1/3rd power in octave band filtered output as features to train K-Nearest Neighbor (KNN) and support vector machine (SVM) classifiers. The SVM classifier with the quadratic kernel is able to identify singletons and twins with a positive predictive value of 100% and 79.1% respectively. The KNN classifier with k=10 neighbors is able to identify singletons and twins with a positive predictive value of 100% and 81.8% respectively.Clinical Relevance: Identifying twin pregnancies from singleton is an essential clinical protocol followed during late pregnancy as there may be complications like twin-twin transfusion syndrome, selective fetal growth restriction, and preterm labor in twin pregnancy [1], [2]. Ultrasound imaging is the most commonly used technique for twin pregnancy detection, though it is often not affordable or available in rural or low-income populations. Utilization of fPCG in such circumstances has immense clinical potential.

Normal and Abnormal Classification of Electrocardiogram: A Primary Screening Tool Kit.

Cardiovascular diseases (CVDs) are one of the principal causes of death. Cardiac arrhythmia, a critical CVD, can be easily detected from an electrocardiogram (ECG) recording. Automated ECG analysis can help clinicians to identify arrhythmia and prevent untimely death. This paper presents a simple model to classify the ECG recordings into two classes: Normal and Abnormal based on morphological and heart rate variability (HRV) features. Before feature extraction, Signal quality analysis (SQA) is performed to abandon poor quality ECG signals. Several machine-learning classifiers such as Support Vector Machine (SVM), Adaboost (AB), Random Forest (RF), Extra-Tree Classifier (ET), Decision Tree (DT), Artificial Neural Network (ANN), K-Nearest Neighbors (KNN), Logistic Regression (LR), Naïve Bayes (NB), and Gradient Boosting (GB) are explored on the extracted feature space. To enhance the study, few feature selection algorithms such as F test, Least Absolute Shrinkage and Selection Operator (LASSO), and Minimal Redundancy Maximal Relevance (mRMR) algorithms are also applied and the outcomes of each algorithm along with the considered classifiers are analyzed and compared. The proposed algorithm is validated on 2648 Normal and 2518 Abnormal ECG recordings. The accuracy of our best classifier is found to be 95.25 %. It is anticipated that the proposed model will be helpful as a primary and mass screening tool kit in clinical settings.