QRS complex detection based on continuous density hidden Markov models using univariate observations

The automatic analysis of the electrocardiogram (ECG) begins, traditionally, with the detection of QRS complexes. Afterwards, useful information can be extracted from it, ranging from the estimation of the instantaneous heart rate to nonlinear heart rate variability analysis. A plethora of works have been published on this topic; consequently, there exist many QRS complex detectors with high-performance values. However, just a few detectors have been conceived that profit from the information contained in several ECG leads to provide a robust QRS complex detection. In this work, we explore the fusion of multi-channel ECG recordings QRS detections as a means to improve the detection performance. This paper presents a decentralized multi-channel QRS complex fusion scheme that optimally combines single-channel detections to produce a single detection signal. Using six different widely used QRS complex detectors on the MIT-BIH Arrhythmia and INCART databases, a reduction in false and missed detections was achieved with the proposed approach compared with the single-channel counterpart. Furthermore, our detection results are comparable with the performance of other multi-channel detectors found in the literature, showing, in turn, various advantages in scalability, adaptability, and simplicity in the system's implementationGraphical AbstractN QRS complex detectors simultaneously monitor N ECG channels. Once a detection occurs in a given channel, a 150 ms long window is opened to look for detections in other channels. Within this window, yn = + 1 if a QRS complex is detected and yn = - 1 otherwise. A coefficient α n, obtained during a training period and related to the detection performance in channel n, multiplies the detection signal yn, so that greater weights are assigned to ECG channels where single-channel detectors performed better. Finally, the binary detection decision (f ) is obtained from the comparison of the weighted sum of single-channel detections (z) with a fixed threshold (β).

Heart rate variability (HRV) analysis is used as a marker of autonomic nervous system activity which may be related to mental and/or physical activity. HRV features can be extracted by detecting QRS complexes from an electrocardiogram (ECG) signal. The difficulties in QRS complex detection are due to the artifacts and noises that may appear in the ECG signal when subjects are performing their daily life activities such as exercise, posture changes, climbing stairs, walking, running, etc. This study describes a strong computation method for real-time QRS complex detection. The detection is improved by the prediction of the position of [Formula: see text] waves by the estimation of the RR intervals lengths. The estimation is done by computing the intensity of the electromyogram noises that appear in the ECG signals and known here in this paper as ECG Trunk Muscles Signals Amplitude (ECG-TMSA). The heart rate (HR) and ECG-TMSA increases with the movement of the subject. We use this property to estimate the lengths of the RR intervals. The method was tested using famous databases, and also with signals acquired when an experiment with 17 subjects from our laboratory. The obtained results using ECG signals from the MIT-Noise Stress Test Database show a QRS complex detection error rate (ER) of 9.06%, a sensitivity of 95.18% and a positive prediction of 95.23%. This method was also tested against MIT-BIH Arrhythmia Database, the result are 99.68% of sensitivity and 99.89% of positive predictivity, with ER of 0.40%. When applied to the signals obtained from the 17 subjects, the algorithm gave an interesting result of 0.00025% as ER, 99.97% as sensitivity and 99.99% as positive predictivity.

This paper introduces an improved HMM (hidden Markov model) for low altitude acoustic target recognition. To overcome the limitation of the classical CDHMM (continuous density hidden Markov model) training algorithm and the generalization ability deficiency of existing discriminative learning methods, a new discriminative training method for estimating the CDHMM in acoustic target recognition is proposed based on the principle of maximizing the minimum relative separation margin. According to the definition of the relative margin, the new training criterion can be equation as a standard constrained minimax optimization problem. Then, the optimization problem can be solved by a GPD (generalized probabilistic descent) algorithm. The experimental results show that the performance of the algorithm is significantly improved compared with the former training method, which can effectively improve the recognition ability of the acoustic target recognition system.

An investigation into the use of Bayesian learning of the parameters of a multivariate Gaussian mixture density has been carried out. In a continuous density hidden Markov model (CDHMM) framework, Bayesian learning serves as a unified approach for parameter smoothing, speaker adaptation, speaker clustering, and corrective training. The goal of this study is to enhance model robustness in a CDHMM-based speech recognition system so as to improve performance. Our approach is to use Bayesian learning to incorporate prior knowledge into the CDHMM training process in the form of prior densities of the HMM parameters. The theoretical basis for this procedure is presented and preliminary results applying to HMM parameter smoothing, speaker adaptation, and speaker clustering are given.Performance improvements were observed on tests using the DARPA RM task. For speaker adaptation, under a supervised learning mode with 2 minutes of speaker-specific training data, a 31% reduction in word error rate was obtained compared to speaker-independent results. Using Baysesian learning for HMM parameter smoothing and sex-dependent modeling, a 21% error reduction was observed on the FEB91 test.

Continuous density hidden Markov models (CD-HMMs) are an essential component of modern systems for automatic speech recognition (ASR). These models assign probabilities to the sequences of acoustic feature vectors extracted by signal processing of speech waveforms. In this chapter, we investigate a new framework for parameter estimation in CD-HMMs. Our framework is inspired by recent parallel trends in the fields of ASR and machine learning. In ASR, significant improvements in performance have been obtained by discriminative training of acoustic models. In machine learning, significant improvements in performance have been obtained by discriminative training of large margin classifiers. Building on both these lines of work, we show how to train CD-HMMs by maximizing an appropriately defined margin between correct and incorrect decodings of speech waveforms. We start by defining an objective function over a transformed parameter space for CD-HMMs, then describe how it can be optimized efficiently by simple gradient-based methods. Within this framework, we obtain highly competitive results for phonetic recognition on the TIMIT speech corpus. We also compare our framework for large margin training to other popular frameworks for discriminative training of CD-HMMs.

Studies on retrospective electrocardiogram (ECG) recordings of patients during cardiac arrest have shown significant changes in heart rate variability (HRV) indices prior to the onset of cardiac arrhythmia. The early detection of these changes in HRV indices increases the chances for a successful medical intervention by increasing the response time window. A portable, handheld remote ECG monitor designed in this research detects the QRS complex and calculates short-term HRV indices in real-time. The QRS detection of the ECG recordings of subjects from the MIT-Arrhythmia database yielded a mean sensitivity of 99.34% and a specificity of 99.31%. ECG recordings from normal subjects and subjects with congestive heart failure were used to identify the differences in HRV indices. An increase in heart rate, high-frequency spectral power (HFP), total spectral power, the ratio of HFP to low-frequency spectral power (LFP), and a decrease in root mean square sum of RR differences were observed. No difference was found on comparison of the standard deviation of normal to normal interval between adjacent R-waves, LFP, and very-low-frequency spectral power. Based on these, additional analytical calculations could be made to provide early warnings of impending cardiac conditions.

BackgroundThe QRS complex is the most noticeable feature in the electrocardiogram (ECG) signal, therefore, its detection is critical for ECG signal analysis. The existing detection methods largely depend on hand-crafted manual features and parameters, which may introduce significant computational complexity, especially in the transform domains. In addition, fixed features and parameters are not suitable for detecting various kinds of QRS complexes under different circumstances.MethodsIn this study, based on 1-D convolutional neural network (CNN), an accurate method for QRS complex detection is proposed. The CNN consists of object-level and part-level CNNs for extracting different grained ECG morphological features automatically. All the extracted morphological features are used by multi-layer perceptron (MLP) for QRS complex detection. Additionally, a simple ECG signal preprocessing technique which only contains difference operation in temporal domain is adopted.ResultsBased on the MIT-BIH arrhythmia (MIT-BIH-AR) database, the proposed detection method achieves overall sensitivity Sen = 99.77%, positive predictivity rate PPR = 99.91%, and detection error rate DER = 0.32%. In addition, the performance variation is performed according to different signal-to-noise ratio (SNR) values.ConclusionsAn automatic QRS detection method using two-level 1-D CNN and simple signal preprocessing technique is proposed for QRS complex detection. Compared with the state-of-the-art QRS complex detection approaches, experimental results show that the proposed method acquires comparable accuracy.

This paper is concerned with the segmentation of the second heart sound (S2) of the phonocardiogram (PCG), in its two acoustic events, aortic (A2) and pulmonary (P2) components. The aortic valve (A2) usually closes before the pulmonary valve (P2) and the delay between these two events is known as 'split' and is typically less than 30 miliseconds. S2 splitting, reverse splitting or reverse occurrence of components A2 and P2 are the most important aspects regarding cardiac diagnosis carried out by the analysis of S2 cardiac sound. An automatic technique, based on discrete wavelet transform and hidden Markov models, is proposed in this paper to segment S2, to estimate de order of occurrence of A2 and P2 and finally to estimate the delay between these two components (split). A discrete density hidden Markov model (DDHMM) is used for phonocardiogram segmentation while embedded continuous density hidden Markov models are used for acoustic models, which allows segmenting S2. Experimental results were evaluated on data collected from five different subjects, using CardioLab system and a Dash family patient monitor. The ECG leads I, II and III and an electronic stethoscope signal were sampled at 977 samples per second.

This paper presents three novel methods for speaker identification of which two methods utilize both the continuous density hidden Markov model (HMM) and the generalized fuzzy model (GFM), which has the advantages of both Mamdani and Takagi-Sugeno models. In the first method, the HMM is utilized for the extraction of shape-based batch feature vector that is fitted with the GFM to identify the speaker. On the other hand, the second method makes use of the Gaussian mixture model (GMM) and the GFM for the identification of speakers. Finally, the third method has been inspired by the way humans cash in on the mutual acquaintances while identifying a speaker. To see the validity of the proposed models [HMM-GFM, GMM-GFM, and HMM-GFM (fusion)] in a real-life scenario, they are tested on VoxForge speech corpus and on the subset of the 2003 National Institute of Standards and Technology evaluation data set. These models are also evaluated on the corrupted VoxForge speech corpus by mixing with different types of noisy signals at different values of signal-to-noise ratios, and their performance is found superior to that of the well-known models.

Measurement and Rate Correction of the QT Interval

This paper describes an automatic text‐ and speaker‐independent language identification method based on hidden Markov models (HMMs) for acoustic features. The hidden Markov modeling is used to represent the phonotactics for each language. Each language has its proper phonotactics.The HMM topology here is a fully structured (ergodic) model wherein any state could transit to all states. Two kinds of HMMs are used: the discrete HMM (DHMM) with the codebook and the continuous density HMM (CHMM). The HMM was trained using both the Baum‐Welch (forward‐backward) algorithm and the Viterbi algorithm. The latter was used for emphasizing the state transition probability.For comparison, experiments also were conducted on the identification using a mixtured Gaussian distribution model with one state. This single‐state Gaussian distribution model gave the same performance as the HMM trained with the Baum‐Welch algorithm. This is because the transition between states which reflects the characteristics of each language does not affect the likelihood scores very much. This problem was addressed by emphasizing the transition probabilities and using the Viterbi algorithm, which resulted in an improvement in the recognition rates. The trigram for optimal state sequence is introduced. Combining it with the HMM produced the best results.

Blind separation of non-stationary sources using continuous density hidden Markov models

Background: The QRS pattern on the electrocardiogram (ECG) has an expected progressively increasing amplitude of the R waves in the anterior leads. Poor R wave progression (low R wave amplitude of < 3 mm in V3) is commonly reported, however its clinical significance is unknown. It may be related to underlying cardiac pathology, body habitus, metabolic abnormalities or ECG lead placement errors. Studying the prevalence and significance of these findings in different genders and races will also shed some light on their clinical utility and applicability. Our objective is to assess the epidemiology and clinical significance of “Poor or Abnormal R-Wave Progression” as reported on routine ECGs. Methods: A database of ECGs performed at the University of South Alabama Medical Center ECG database (Intellispace and/or Cerner) was downloaded and compiled into searchable PDF format. ECGs were searched for patients with “Abnormal or Poor R Wave Progression”. Available clinical data on patients with these ECG findings were analyzed and reported as absolute numbers and percentages. Results: The total number of ECG with “Poor R-Wave Progression” were 318. Out of these, 209 (66%) were females and 109 (34%) males; 209 (66%) were black, 105 (33%) were white and 4 (1%) other races. There were 143 (45%) black females, 65 (20%) white females, 66 (21%) black males and 40 (13%) white males. There were 234 (74%) patients < 60 years of age while 85 (26%) were 60 years of age or older. There were 233 (73%) with hypertension. Only 198 (62%) had echocardiograms, with an average ejection fraction (EF) of 52%. Out of these 49 ((25%) had mild to moderate decrease in EF ( > 30 - 50%) while 28 (14%) had severely decreased EF of 30% or less. There were 81 (41%) with wall motion abnormalities concerning for coronary disease, while 95 (48%) had diastolic dysfunction. Discussion: Poor R-wave progression is a commonly reported finding on ECG. It is more prevalent in females, which may be due to improper lead placement or chest size. However, its higher prevalence in blacks may be indicative of higher incidence of heart failure, coronary disease and hypertension. It is more frequent in younger patients who may be more obese, but also carry multiple risk factors for cardiac disease. Poor R wave progression showed strong correlation with any degree of heart failure and with wall motion abnormalities on echocardiogram concerning for coronary disease. Therefore this finding should trigger a diagnostic cascade with echocardiography and/or stress testing, depending on the clinical scenario, in order to identify potentially serious underlying cardiac pathology which may be amenable to therapy.

An automated patient-specific ECG beat classification using LSTM-based recurrent neural networks

Heart disease is a heart condition that sometimes causes a person to die suddenly. One indication is a rhythm disorder known as arrhythmia. Multi-class Arrhythmia Detection has followed: QRS complex detection procedure and arrhythmia classification based on the QRS complex morphology. We proposed an edge device that detects QRS complexes based on variance analysis (QVAT) and the arrhythmia classification based on the QRS complex spectrogram. The classifier uses two-dimensional convolutional neural network (2D CNN) deep learning. We use a single board computer and neural network compute stick to implement the edge device. The outcomes are a prototype device cardiologists use as a supporting tool for analysing ECG signals, and patients can also use it for self-tests to figure out their heart health. To evaluate the performance of our edge device, we tested using the MIT-BIH database because other methods also use the data. The QVAT sensitivity and predictive positive are 99.81% and 99.90%, respectively. Our classifier's accuracy, sensitivity, predictive positive, specificity, and F1-score are 99.82%, 99.55%, 99.55%, 99.89%, and 99.55%, respectively. The experiment result of arrhythmia classification shows that our method outperforms the others. Still, for r-peak detection, the QVAT implemented in an edge device is comparable to the other methods. In future work, we can improve the performance of r-peak detection using the double-check algorithm in QVAT and cross-check the QRS complex detection by adding 1 class to the classifier, namely the non-QRS class.