This paper presents a signal processing algorithm to remove different types of noise from a gyroscopic head-borne computer mouse. The proposed algorithm is a combination of a Kalman filter (KF), a Weighted-frequency Fourier Linear Combiner (WFLC) and a threshold with delay method (TWD). The gyroscopic head-borne mouse was developed to assist persons with movement disorders. However, since MEMS-gyroscopes are usually sensitive to environmental disturbances such as shock, vibration and temperature change, a large portion of noise is added at the same time as the head movement is sensed by the MEMS-gyroscope. The combined method is applied to the specially adapted mouse, to filter out different types of noise together with the offset and drift, with marginal need of the calculation capacity. The method is examined with both static state tests and movement operation tests. Angular position is used to evaluate the errors. The results demonstrate that the combined method improved the head motion signal substantially, with 100.0% error reduction during the static state, 98.2% position error correction in the case of movements without drift and 99.9% with drift. The proposed combination in this paper improved the static stability and position accuracy of the gyroscopic head-borne mouse system by reducing noise, offset and drift, and also has the potential to be used in other gyroscopic sensor systems to improve the accuracy of signals.
Deep Learning (DL) has recently become a topic of study in different applications including healthcare, in which timely detection of anomalies on Electrocardiogram (ECG) can play a vital role in patient monitoring. This paper presents a comprehensive review study on the recent DL methods applied to the ECG signal for the classification purposes. This study considers various types of the DL methods such as Convolutional Neural Network (CNN), Deep Belief Network (DBN), Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU). From the 75 studies reported within 2017 and 2018, CNN is dominantly observed as the suitable technique for feature extraction, seen in 52% of the studies. DL methods showed high accuracy in correct classification of Atrial Fibrillation (AF) (100%), Supraventricular Ectopic Beats (SVEB) (99.8%), and Ventricular Ectopic Beats (VEB) (99.7%) using the GRU/LSTM, CNN, and LSTM, respectively
This paper proposes a framework for binary classification of the time series with cyclic characteristics. The framework presents an iterative algorithm for learning the cyclic characteristics by introducing the discriminative frequency bands (DFBs) using the discriminant analysis along with k-means clustering method. The DEBs are employed by a hybrid model for learning dynamic characteristics of the time series within the cycles, using statistical and structural machine learning techniques. The framework offers a systematic procedure for finding the optimal design parameters associated with the hybrid model. The proposed model is optimized to detect the changes of the heart sound recordings (HSRs) related to aortic stenosis. Experimental results show that the proposed framework provides efficient tools for the classification of the HSRs based on the heart murmurs. It is also evidenced that the hybrid model, proposed by the framework, substantially improves the classification performance when it comes to detection of the heart disease.
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.
This paper presents an original method for extracting medical information from a heart sound recording, so called Phonocardiographic (PCG) signal. The extracted information is employed by a binary classifier to distinguish between stenosis and regurgitation murmurs. The method is based on using our original neural network, the Time-Growing Neural Network (TGNN), in an innovative way. Children with an obstruction on their semilunar valve are considered as the patient group (PG) against a reference group (RG) of children with a regurgitation in their atrioventricular valve. PCG signals were collected from 55 children, 25/30 from the PG/RG, who referred to the Children Medical Center of Tehran University. The study was conducted according to the guidelines of Good Clinical Practices and the Declaration of Helsinki. Informed consents were obtained for all the patients prior to the data acquisition. The accuracy and sensitivity of the method was estimated to be 85% and 80% respectively, exhibiting a very good performance to be used as a part of decision support system. Such a decision support system can improve the screening accuracy in primary healthcare centers, thanks to the innovative use of TGNN.
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.
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.
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.
This paper presents a novel hybrid model for classifying time series of heart sound signal using time-growing neural network. The proposed hybrid model takes segmental behaviour of heart sound signal into account by combining two different deep learning methods, the Static and the Moving Time-Growing Neural Network, which we call STGNN and MTGNN, respectively. Flexibility of the model in learning both deterministic and stochastic segments of signal allows it to learn those complicated characteristics of heart sound signal caused by any obstruction on semilunar heart valve. The model is trained to distinguish between a patient group and a reference group. The patient group is comprised of the subjects with the semilunar heart valve abnormalities including aortic stenosis, pulmonary stenosis and bicuspid aortic valve, whereas the reference group which is composed of the individuals with the heart abnormalities other than those of the reference group or the healthy ones. The model is validated using two different databases: one comprised of 140 children with various heart conditions, and the other one constituted of 50 elderly patients with aortic stenosis. Both the datasets were collected from the referrals to the University hospitals. The overall accuracy and sensitivity of the model are estimated to be 84.2% and 82.8%, respectively. The results show that the model exhibits sufficient capability to distinguish between the patient and the reference group in such a demanding clinical application.
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.
This paper presents a machine learning method to detect and discriminate between Aortic Stenosis (AS) and Bicuspid Aortic Valve (BAV) based on heart sound analysis. Differentiation between the two heart conditions is clinically important, but complicated if relying merely on the conventional auscultation. A novel form of the Time Growing Neural Network (TGNN) is introduced for the classification purpose. The method is applied to a dataset comprised of 87 children referrals to a university hospital, from which 50 individuals are healthy (with and without innocent murmur), and the rest are abnormal with either AS (15 individuals) or BAV (22 individuals). The baseline for comparison is a Time-Delayed Neural Network (TDNN) with the same size of the feature vector and the temporal frame. We used our original validation methods, named A-Test, which provides valuable information about structural risk and also learning capacity of any supervised classification method. A-Test is an elaborated version of K-Fold validation method, in a rather profound way. Performance of the TGNN is superior comparing to the presented TDNN, with an accuracy of 85.8% against 81.5%. This method can be integrated with our intelligent phonocardiography to serve as an enhanced assessment tool in hands of nurses or practitioners at primary healthcare centers.
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.
This paper presents a novel method for processing heart sound signal for screening forth heart sound (S4). The proposed method is based on time growing neural network with a new scheme, which we call the Backward Time-Growing Neural Network (BTGNN). The BTGNN is trained for detecting S4 in recordings of heart sound signal. In total, 83 children patients, referred to a children University hospital, participated in the study. The collected signals are composed of the subjects with and without S4 for training and testing the method. Performance of the method is evaluated using the Leave-One-Out and the repeated random sub sampling methods. The accuracy/sensitivity of the method is estimated to be 88.3%/82.4% and the structural risk is calculated to be 18.3% using repeated random sub sampling and the A-Test methods, respectively.
The present invention is related to a method for the determination of frequency band characteristics of a heart disease. A first set of phonocardiograms are recorded from a first set of reference healthy patients, and a second set of phonocardiograms from a second set of patients suffering of a heart disease. Spectral energies of all possible frequency bands are then calculated. These spectral energies are then compared in order to determine an optimized frequency band that gives rise to the maximal distinction between spectral energies of the phonocardiograms from first and second set of phonocardiograms.
This study presents an algorithm for screening aortic stenosis, based on heart sound signal processing. It benefits from an artificial intelligent-based (AI-based) model using a multi-layer perceptron neural network. The AI-based model learns disease related murmurs using non-stationary features of the murmurs. Performance of the model is statistically evaluated using two different databases, one of children and the other of elderly volunteers with normal heart condition and aortic stenosis. Results showed a 95% confidence interval of the high accuracy/sensitivity thus exhibiting a superior performance to a cardiologist who relies on the conventional auscultation. The study suggests including the heart sound signal in the clinical decision making due to its potential to improve the screening accuracy.
This study presents a screening algorithm for severe aortic stenosis (AS), based on a processing method for phonocardiographic (PCG) signal. The processing method employs a hybrid model, constituted of a hidden Markov model and support vector machine. The method benefits from a preprocessing phase for an enhanced learning. The performance of the method is statistically evaluated using PCG signals recorded from 50 individuals who were referred to the echocardiography lab at Linköping University hospital. All the individuals were diagnosed as having a degree of AS, from mild to severe, according to the echocardiographic measurements. The patient group consists of 26 individuals with severe AS, and the rest of the 24 patients comprise the control group. Performance of the method is statistically evaluated using repeated random sub sampling. Results showed a 95% confidence interval of (80.5%-82.8%) /(77.8%- 80.8%) for the accuracy/sensitivity, exhibiting an acceptable performance to be used as decision support system in the primary healthcare center.
This paper presents a sophisticated machine learning method for screening children with patent ductus arteriosus (PDA) using the phonocardiogram recording as the input signal. The method is based on our original algorithm for finding the discriminative contents of the signal for the healthy children and the diseased ones, which we called Short Time Arash-Band (STAB) method. The STAB employs our specific discriminant analysis method for finding the joint temporal and spectral characteristics of the signal which provide the optimal segregation. Fifty pediatric referrals to a children university hospital, composed of 30 healthy and 20 with PDA, were participated in this study after obtaining the informed consent, according to the guidelines of the hospital which is in compliance with the declaration of Helsinki. The accuracy/sensitivity of the method was evaluated to be 86%/85%, using the leave-one-out validation method. Results show a potential for the approach in pediatric cardiac assessments that is considered as a demanding clinical application. Such an approach, which we called the intelligent phonocardiography, can be easily employed by the nurses or practitioners to improve the screening accuracy in the primary healthcare centers.
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.
In this paper, we present a novel algorithm for pediatric heart sound segmentation, incorporated into a graphical user interface. The algorithm employs both the Electrocardiogram (ECG) and Phonocardiogram (PCG) signals for an efficient segmentation under pathological circumstances. First, the ECG signal is invoked in order to determine the beginning and end points of each cardiac cycle by using wavelet transform technique. Then, first and second heart sounds within the cycles are identified over the PCG signal by paying attention to the spectral properties of thesounds. The algorithm is applied on 120 recordings of normal and pathological children, totally containing 1976 cardiac cycles. The accuracy of thesegmentation algorithm is 97% for S1 and 94% for S2 identification while all the cardiac cycles are correctly determined.
This paper suggests an internet-based tool for cardiac diagnosis in children. The main focus of the paper is the intelligent algorithms for processing heart sounds that are implementable on an internet platform. The algorithms are based on the statistical classification methods, tailored for the heart sound signal processing. The algorithms, applied to 55 healthy and 45 children with congenital heart diseases. The accuracy of the algorithm is estimated to be 86.0 % in screening the children with pathological murmurs, and 95.7 %, 92.9 % and 91.4 % in detecting the children with aortic stenosis, pulmonary stenosis and mitral regurgitation, respectively, showing an acceptable performance to be employed as a decision support tool.
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. © 2017 European Federation for Medical Informatics (EFMI) and IOS Press.
This study presents an artificial intelligent-based method for processing phonocardiographic (PCG) signal of the patients with ejection murmur to assess the underlying pathology initiating the murmur. The method is based on our unique method for finding disease-related frequency bands in conjunction with a sophisticated statistical classifier. Children with aortic stenosis (AS), and pulmonary stenosis (PS) were the two patient groups subjected to the study, taking the healthy ones (no murmur) as the control group. PCG signals were acquired from 45 referrals to the children University hospital, comprised of 15 individuals of each group; all were diagnosed by the expert pediatric cardiologists according to the echocardiographic measurements together with the complementary tests. The accuracy of the method is evaluated to be 90% and 93.3% using the 5-fold and leave-one-out validation method, respectively. The accuracy is slightly degraded to 86.7% and 93.3% when a Gaussian noise with signal to noise ratio of 20 dB is added to the PCG signals, exhibiting an acceptable immunity against the noise. The method offered promising results to be used as a decision support system in the primary healthcare centers or clinics.
This paper presents a novel method for detecting cardiac ejection murmurs from other pathological and physiological heart murmurs in children. The proposed method combines a hybrid model and a time growing neural network for an improved detection even in mild condition. Children with aortic stenosis and pulmonary stenosis comprised the patient category against the reference category containing mitral regurgitation, ventricular septal defect, innocent murmur and normal (no murmur) conditions. In total, 120 referrals to a children University hospital participated to the study after giving their informed consent. Confidence interval of the accuracy, sensitivity and specificity is estimated to be 87.2% ̶ 88.8%, 83.4% ̶ 86.9% and 88.3% ̶ 90.0%, respectively.
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. .
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.
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 (Si) 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 Si and S2 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.
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.