Accurate detection of paroxysmal atrial fibrillation with certified-GAN and neural architecture search.

This paper presents a novel machine learning framework for detecting PxAF, a pathological characteristic of electrocardiogram (ECG) that can lead to fatal conditions such as heart attack. To enhance the learning process, the framework involves a generative adversarial network (GAN) along with a neural architecture search (NAS) in the data preparation and classifier optimization phases. The GAN is innovatively invoked to overcome the class imbalance of the training data by producing the synthetic ECG for PxAF class in a certified manner. The effect of the certified GAN is statistically validated. Instead of using a general-purpose classifier, the NAS automatically designs a highly accurate convolutional neural network architecture customized for the PxAF classification task. Experimental results show that the accuracy of the proposed framework exhibits a high value of 99.0% which not only enhances state-of-the-art by up to 5.1%, but also improves the classification performance of the two widely-accepted baseline methods, ResNet-18, and Auto-Sklearn, by [Formula: see text] and [Formula: see text].

Recurrent vs Non-Recurrent Convolutional Neural Networks for Heart Sound Classification.

Convolutional Neural Network (CNN) has been widely proposed for different tasks of heart sound analysis. This paper presents the results of a novel study on the performance of a conventional CNN in comparison to the different architectures of recurrent neural networks combined with CNN for the classification task of abnormal-normal heart sounds. The study considers various combinations of parallel and cascaded integration of CNN with Gated Recurrent Network (GRN) as well as Long- Short Term Memory (LSTM) and explores the accuracy and sensitivity of each integration independently, using the Physionet dataset of heart sound recordings. The accuracy of the parallel architecture of LSTM-CNN reached 98.0% outperforming all the combined architectures, with a sensitivity of 87.2%. The conventional CNN offered sensitivity/accuracy of 95.9%/97.3% with far less complexity. Results show that a conventional CNN can appropriately perform and solely employed for the classification of heart sound signals.

Parralel Recurrent Convolutional Neural Network for Abnormal Heart Sound Classification.

This paper presents the results of a study performed on Parallel Convolutional Neural Network (PCNN) toward detecting heart abnormalities from the heart sound signals. The PCNN preserves dynamic contents of the signal in a parallel combination of the recurrent neural network and a Convolutional Neural Network (CNN). The performance of the PCNN is evaluated and compared to the one obtained from a Serial form of the Convolutional Neural Network (SCNN) as well as two other baseline studies: a Long- and Short-Term Memory (LSTM) neural network and a Conventional CNN (CCNN). We employed a well-known public dataset of heart sound signals: the Physionet heart sound. The accuracy of the PCNN, was estimated to be 87.2% which outperforms the rest of the three methods: the SCNN, the LSTM, and the CCNN by 12%, 7%, and 0.5%, respectively. The resulting method can be easily implemented in an Internet of Things platform to be employed as a decision support system for the screening heart abnormalities.

Deep Time Growing Neural Network vs Convolutional Neural Network for Intelligent Phonocardiography.

This paper explores the capabilities of a sophisticated deep learning method, named Deep Time Growing Neural Network (DTGNN), and compares its possibilities against a generally well-known method, Convolutional Neural network (CNN). The comparison is performed by using time series of the heart sound signal, so-called Phonocardiography (PCG). The classification objective is to discriminate between healthy and patients with cardiac diseases by applying a deep machine learning method to PCGs. This approach which is called intelligent phonocardiography has received interest from the researchers toward the development of a smart stethoscope for decentralized diagnosis of heart disease. It is found that DTGNN associates further flexibility to the approach which enables the classifier to learn subtle contents of PCG, and meanwhile better copes with the complexities intrinsically that exist in the medical applications such as the imbalance training. The structural risk of the two methods is compared using the A-Test method.

A-Test Method for Quantifying Structural Risk and Learning Capacity of Supervised Machine Learning Methods.

This paper presents an original method for studying the performance of the supervised Machine Learning (ML) methods, the A-Test method. The method offers the possibility of investigating the structural risk as well as the learning capacity of ML methods in a quantitating manner. A-Test provides a powerful validation method for the learning methods with small or medium size of the learning data, where overfitting is regarded as a common problem of learning. Such a condition can occur in many applications of bioinformatics and biomedical engineering in which access to a large dataset is a challengeable task. Performance of the A-Test method is explored by validation of two ML methods, using real datasets of heart sound signals. The datasets comprise of children cases with a normal heart condition as well as 4 pathological cases: aortic stenosis, ventricular septal defect, mitral regurgitation, and pulmonary stenosis. It is observed that the A-Test method provides further comprehensive and more realistic information about the performance of the classification methods as compared to the existing alternatives, the K-fold validation and repeated random sub-sampling.

Distinguishing Septal Heart Defects from the Valvular Regurgitation Using Intelligent Phonocardiography.

This paper presents an original machine learning method for extracting diagnostic medical information from heart sound recordings. The method is proposed to be integrated with an intelligent phonocardiography in order to enhance diagnostic value of this technology. The method is tailored to diagnose children with heart septal defects, the pathological condition which can bring irreversible and sometimes fatal consequences to the children. The study includes 115 children referrals to an university hospital, consisting of 6 groups of the individuals: atrial septal defects (10), healthy children with innocent murmur (25), healthy children without any murmur (25), mitral regurgitation (15), tricuspid regurgitation (15), and ventricular septal defect (25). The method is trained to detect the atrial or ventricular septal defects versus the rest of the groups. Accuracy/sensitivity and the structural risk of the method is estimated to be 91.6%/88.4% and 9.89%, using the repeated random sub sampling and the A-Test method, respectively.

An Edge Computing Method for Extracting Pathological Information from Phonocardiogram.

This paper presents a structure of decision support system for pediatric cardiac disease, based on an Internet of Things (IoT) framework. The structure performs the intelligent decision making at its edge processing level, which classifies the heart sound signal, to three classes of cardiac conditions, normal, mild disease, and critical disease. Three types of the errors are introduced to evaluate the performance of the processing method, Type 1, 2 and 3, defined as the incorrect classification from the critical disease, mild, and normal, respectively. The method is validated using 140 real data patient records collected from the hospital referrals. The estimated negative errors for the Type 1, and 2, are calculated to be 0% and 4.8%, against the positive errors which are 6.3% and 13.3%, respectively. The Type 3, is calculated to be 16.7%, showing a high sensitivity of the method to be used in an IoT healthcare framework.

Structural Risk Evaluation of a Deep Neural Network and a Markov Model in Extracting Medical Information from Phonocardiography.

This paper presents a method for exploring structural risk of any artificial intelligence-based method in bioinformatics, the A-Test method. This method provides a way to not only quantitate the structural risk associated with a classification method, but provides a graphical representation to compare the learning capacity of different classification methods. Two different methods, Deep Time Growing Neural Network (DTGNN) and Hidden Markov Model (HMM), are selected as two classification methods for comparison. Time series of heart sound signals are employed as the case study where the classifiers are trained to learn the disease-related changes. Results showed that the DTGNN offers a superior performance both in terms of the capacity and the structural risk. The A-Test method can be especially employed in comparing the learning methods with small data size.

A Deep Machine Learning Method for Classifying Cyclic Time Series of Biological Signals Using Time-Growing Neural Network.

This paper presents a novel method for learning the cyclic contents of stochastic time series: the deep time-growing neural network (DTGNN). The DTGNN combines supervised and unsupervised methods in different levels of learning for an enhanced performance. It is employed by a multiscale learning structure to classify cyclic time series (CTS), in which the dynamic contents of the time series are preserved in an efficient manner. This paper suggests a systematic procedure for finding the design parameter of the classification method for a one-versus-multiple class application. A novel validation method is also suggested for evaluating the structural risk, both in a quantitative and a qualitative manner. The effect of the DTGNN on the performance of the classifier is statistically validated through the repeated random subsampling using different sets of CTS, from different medical applications. The validation involves four medical databases, comprised of 108 recordings of the electroencephalogram signal, 90 recordings of the electromyogram signal, 130 recordings of the heart sound signal, and 50 recordings of the respiratory sound signal. Results of the statistical validations show that the DTGNN significantly improves the performance of the classification and also exhibits an optimal structural risk.

Intelligent Phonocardiography for Screening Ventricular Septal Defect Using Time Growing Neural Network.

This paper presents results of a study on the applicability of the intelligent phonocardiography in discriminating between Ventricular Spetal Defect (VSD) and regurgitation of the atrioventricular valves. An original machine learning method, based on the Time Growing Neural Network (TGNN), is employed for classifying the phonocardiographic recordings collected from the pediatric referrals to a children hospital. 90 individuals, 30 VSD, 30 with the valvular regurgitation, and 30 healthy subjects, participated in the study after obtaining the informed consents. The accuracy and sensitivity of the approach is estimated to be 86.7% and 83.3%, respectively, showing a good performance to be used as a decision support system.

A Decision Support System for Cardiac Disease Diagnosis Based on Machine Learning Methods.

This paper proposes a decision support system for screening pediatric cardiac disease in primary healthcare centres relying on the heart sound time series analysis. The proposed system employs our processing method which is based on the hidden Markov model for extracting appropriate information from the time series. The binary output resulting from the method is discriminative for the two classes of time series existing in our databank, corresponding to the children with heart disease and the healthy ones. A total 90 children referrals to a university hospital, constituting of 55 healthy and 35 children with congenital heart disease, were enrolled into the study after obtaining the informed consent. Accuracy and sensitivity of the method was estimated to be 86.4% and 85.6%, respectively, showing a superior performance than what a paediatric cardiologist could achieve performing auscultation. The method can be easily implemented using mobile and web technology to develop an easy-to-use tool for paediatric cardiac disease diagnosis.

An Intelligent Phonocardiography for Automated Screening of Pediatric Heart Diseases.

This paper presents a robust device for automated screening of pediatric heart diseases based on our unique processing method in murmur characterization; the Arash-Band method. The present study modifies the Arash-Band method and employs output of the modified method in conjunction with the two other original techniques to extract indicative feature vectors for the screening. The extracted feature vectors are classified by using the support vector machine method. Results show that the proposed modifications significantly enhances performance of the Arash-Band in terms of the both accuracy and sensitivity as the corresponding effect sizes are sufficiently large. The proposed algorithm has been incorporated into an Android-based tablet to constitute an intelligent phonocardiogram with the automatic screening capability. In order to obtain confidence interval of the accuracy and sensitivity, an inferable statistical test is applied on our database containing the phonocardiogram signals recorded from 263 of the referrals to a hospital. The expected value of the accuracy/sensitivity is estimated to be 87.45 % / 87.29 % with a 95 % confidence interval of (80.19 % - 92.47 %) / (76.01 % - 95.78 %) exhibiting superior performance than a pediatric cardiologist who relies on conventional or even computer-assisted auscultation.

A Novel Method for Screening Children with Isolated Bicuspid Aortic Valve.

This paper presents a novel processing method for heart sound signal: the statistical time growing neural network (STGNN). The STGNN performs a robust classification by merging supervised and unsupervised statistical methods to overcome non-stationary behavior of the signal. By combining available preprocessing and segmentation techniques and the STGNN classifier, we build an automatic tool for screening children with isolated BAV, the congenital heart malformation which can lead to serious cardiovascular lesions. Children with BAV (22 individuals) and healthy condition (28 individuals) are subjected to the study. The performance of the STGNN is compared to that of a time growing neural network (CTGNN) and a conventional support vector (CSVM) machine, using balanced repeated random sub sampling. The average of the accuracy/sensitivity is estimated to be 87.4/86.5 for the STGNN, 81.8/83.4 for the CTGNN, and 72.9/66.8 for the CSVM. Results show that the STGNN offers better performance and provides more immunity to the background noise as compared to the CTGNN and CSVM. The method is implementable in a computer system to be employed in primary healthcare centers to improve the screening accuracy.

A novel method for discrimination between innocent and pathological heart murmurs.

This paper presents a novel method for discrimination between innocent and pathological murmurs using the growing time support vector machine (GTSVM). The proposed method is tailored for characterizing innocent murmurs (IM) by putting more emphasis on the early parts of the signal as IMs are often heard in early systolic phase. Individuals with mild to severe aortic stenosis (AS) and IM are the two groups subjected to analysis, taking the normal individuals with no murmur (NM) as the control group. The AS is selected due to the similarity of its murmur to IM, particularly in mild cases. To investigate the effect of the growing time windows, the performance of the GTSVM is compared to that of a conventional support vector machine (SVM), using repeated random sub-sampling method. The mean value of the classification rate/sensitivity is found to be 88%/86% for the GTSVM and 84%/83% for the SVM. The statistical evaluations show that the GTSVM significantly improves performance of the classification as compared to the SVM.

Detection of systolic ejection click using time growing neural network.

In this paper, we present a novel neural network for classification of short-duration heart sounds: the time growing neural network (TGNN). The input to the network is the spectral power in adjacent frequency bands as computed in time windows of growing length. Children with heart systolic ejection click (SEC) and normal children are the two groups subjected to analysis. The performance of the TGNN is compared to that of a time delay neural network (TDNN) and a multi-layer perceptron (MLP), using training and test datasets of similar sizes with a total of 614 normal and abnormal cardiac cycles. From the test dataset, the classification rate/sensitivity is found to be 97.0%/98.1% for the TGNN, 85.1%/76.4% for the TDNN, and 92.7%/85.7% for the MLP. The results show that the TGNN performs better than do TDNN and MLP when frequency band power is used as classifier input. The performance of TGNN is also found to exhibit better immunity to noise.

A novel method for pediatric heart sound segmentation without using the ECG.

In this paper, we propose a novel method for pediatric heart sounds segmentation by paying special attention to the physiological effects of respiration on pediatric heart sounds. The segmentation is accomplished in three steps. First, the envelope of a heart sounds signal is obtained with emphasis on the first heart sound (S(1)) and the second heart sound (S(2)) by using short time spectral energy and autoregressive (AR) parameters of the signal. Then, the basic heart sounds are extracted taking into account the repetitive and spectral characteristics of S(1) and S(2) sounds by using a Multi-Layer Perceptron (MLP) neural network classifier. In the final step, by considering the diastolic and systolic intervals variations due to the effect of a child's respiration, a complete and precise heart sounds end-pointing and segmentation is achieved.

Computerized screening of children congenital heart diseases.

In this paper, we propose a method for automated screening of congenital heart diseases in children through heart sound analysis techniques. Our method relies on categorizing the pathological murmurs based on the heart sections initiating them. We show that these pathelogical murmur categories can be identified by examining the heart sound energy over specific frequency bands, which we call, Arash-Bands. To specify the Arash-Band for a category, we evaluate the energy of the heart sound over all possible frequency bands. The Arash-Band is the frequency band that provides the lowest error in clustering the instances of that category against the normal ones. The energy content of the Arash-Bands for different categories constitue a feature vector that is suitable for classification using a neural network. In order to train, and to evaluate the performance of the proposed method, we use a training data-bank, as well as a test data-bank, collectively consisting of ninety samples (normal and abnormal). Our results show that in more than 94% of cases, our method correctly identifies children with congenital heart diseases. This percentage improves to 100%, when we use the Jack-Knife validation method over all the 90 samples.