Identification of Pulmonary Hypertension Using Entropy Measure Analysis of Heart Sound Signal

Study of the Relationship between Pulmonary Artery Pressure and Heart Valve Vibration Sound Based on Mock Loop

The vibration of the heart valves' closure is an important component of the heart sound and contains important information about the mechanical activity of a heart. Stenosis of the distal pulmonary artery can lead to pulmonary hypertension (PH). Therefore, in this paper, the relationship between the vibration sound of heart valves and the pulmonary artery blood pressure was investigated to contribute to the noninvasive detection of PH. In this paper, a lumped parameter circuit platform of pulmonary circulation was first set to guide the establishment of a mock loop of circulation. By adjusting the distal vascular resistance of the pulmonary artery, six different pulmonary arterial pressure states were achieved. In the experiment, pulmonary artery blood pressure, right ventricular blood pressure, and the vibration sound of the pulmonary valve and tricuspid valve were measured synchronously. Features of the time domain and frequency domain of two valves' vibration sound were extracted. By conducting a significance analysis of the inter-group features, it was found that the amplitude, energy and frequency features of vibration sounds changed significantly. Finally, the continuously varied pulmonary arterial blood pressure and valves' vibration sound were obtained by continuously adjusting the resistance of the distal pulmonary artery. A backward propagation neural network and deep learning model were used, respectively, to estimate the features of pulmonary arterial blood pressure, pulmonary artery systolic blood pressure, the maximum rising rate of pulmonary artery blood pressure and the maximum falling rate of pulmonary artery blood pressure by the vibration sound of the pulmonary and tricuspid valves. The results showed that the pulmonary artery pressure parameters can be well estimated by valve vibration sounds.

Phonocardiogram transfer learning-based CatBoost model for diastolic dysfunction identification using multiple domain-specific deep feature fusion

Left ventricular diastolic dyfunction detection is particularly important in cardiac function screening. This paper proposed a phonocardiogram (PCG) transfer learning-based CatBoost model to detect diastolic dysfunction noninvasively. The Short-Time Fourier Transform (STFT), Mel Frequency Cepstral Coefficients (MFCCs), S-transform and gammatonegram were utilized to perform four different representations of spectrograms for learning the representative patterns of PCG signals in two-dimensional image modality. Then, four pre-trained convolutional neural networks (CNNs) such as VGG16, Xception, ResNet50 and InceptionResNetv2 were employed to extract multiple domain-specific deep features from PCG spectrograms using transfer learning, respectively. Further, principal component analysis and linear discriminant analysis (LDA) were applied to different feature subsets, respectively, and then these different selected features are fused and fed into CatBoost for classification and performance comparison. Finally, three typical machine learning classifiers such as multilayer perceptron, support vector machine and random forest were employed to compared with CatBoost. The hyperparameter optimization of the investigated models was determined through grid search. The visualized result of the global feature importance showed that deep features extracted from gammatonegram by ResNet50 contributed most to classification. Overall, the proposed multiple domain-specific feature fusion based CatBoost model with LDA achieved the best performance with an area under the curve of 0.911, accuracy of 0.882, sensitivity of 0.821, specificity of 0.927, F1-score of 0.892 on the testing set. The PCG transfer learning-based model developed in this study could aid in diastolic dysfunction detection and could contribute to non-invasive evaluation of diastolic function.

Simulation of Acute Pulmonary Hypertension in Beagle Dogs

Acoustic cardiography (AC) combined with heart sound (HS) recording and electrocardiogram (ECG) provides a noninvasive and inexpensive way to understand the electrical mechanical activity of the heart. Pulmonary artery stenosis can cause hemodynamic abnormalities that might lead to pulmonary hypertension (PH). In this paper, we examined the relationships between the acoustic characteristics of the AC and hemodynamic changes in a beagle dog model of PH.Four healthy beagle dogs were injected with the prostaglandin endoperoxide receptor agonist U-44069 to induce acute PH states. AC was employed to analyze the process of pre-PH, intra-PH, and post-PH. Right ventricular blood pressure (RVBP) was measured via right cardiac catheterization, an invasive method performed in parallel for comparative hemodynamic evaluation. As RVBP increased or decreased, the HS features changed accordingly during acute PH occurrence and development. Right ventricular systolic blood pressure (RVSBP) significantly correlated with the minimum of the first HS (S1) amplitude (correlation coefficient (CC) = -0.82), energy of the S1 (CC = 0.86), energy of the second HS (S2) (CC = 0.67), entropy of the S1 (CC = -0.94), and ratio of electromechanical systolic time (EMST) to the cardiac cycle time (CC = 0.81). The two techniques (AC [HSs and ECG] versus right cardiac catheterization [RVBP]) were significantly correlated. Especially, the diastolic filling time (DFT) had a significant relationship with the right ventricular diastolic time (RVDT) (CC = 0.97), perfusion time (PT) (CC = 0.96), and cardiac cycle time (RR) (CC = 0.96). The CCs between the RVDT and the max dp/dt to min dp/dt, the EMST and the Q to min dp/dt, and the electromechanical activation time and the Q to max dp/dt were 0.95, 0.99, and 0.86, respectively. Furthermore, the logistic regression model with different combinations was used to identify the effective features for monitoring hemodynamic and pathophysiologic conditions.AC provided significant insight into mechanical dysfunction in a rapid and noninvasive way that could be used for early screening of PH.

A multi-scale and multi-domain heart sound feature-based machine learning model for ACC/AHA heart failure stage classification

OBJECTIVE

Heart sounds can reflect detrimental changes in cardiac mechanical activity that are common pathological characteristics of chronic heart failure (CHF). The ACC/AHA heart failure (HF) stage classification is essential for clinical decision-making and the management of CHF. Herein, a machine learning model that makes use of multi-scale and multi-domain heart sound features was proposed to provide an objective aid for ACC/AHA HF stage classification.

APPROACH

A dataset containing phonocardiogram (PCG) signals from 275 subjects was obtained from two medical institutions and used in this study. Complementary ensemble empirical mode decomposition and tunable-Q wavelet transform were used to construct self-adaptive sub-sequences and multi-level sub-band signals for PCG signals. Time-domain, frequency-domain and nonlinear feature extraction were then applied to the original PCG signal, heart sound sub-sequences and sub-band signals to construct multi-scale and multi-domain heart sound features. The features selected via the least absolute shrinkage and selection operator were fed into a machine learning classifier for ACC/AHA HF stage classification. Finally, mainstream machine learning classifiers, including least-squares support vector machine (LS-SVM), deep belief network (DBN) and random forest (RF), were compared to determine the optimal model.

MAIN RESULTS

The results showed that the LS-SVM, which utilized a combination of multi-scale and multi-domain features, achieved better classification performance than the DBN and RF using multi-scale or multi-domain features alone or together, with average sensitivity, specificity, and accuracy of 0.821, 0.955 and 0.820 on the testing set, respectively.

SIGNIFICANCE

PCG signal analysis provides efficient measurement information regarding CHF severity and is a promising noninvasive method for ACC/AHA HF stage classification.

Detection of Coronary Artery Disease Using Multi-Domain Feature Fusion of Multi-Channel Heart Sound Signals

Heart sound signals reflect valuable information about heart condition. Previous studies have suggested that the information contained in single-channel heart sound signals can be used to detect coronary artery disease (CAD). But accuracy based on single-channel heart sound signal is not satisfactory. This paper proposed a method based on multi-domain feature fusion of multi-channel heart sound signals, in which entropy features and cross entropy features are also included. A total of 36 subjects enrolled in the data collection, including 21 CAD patients and 15 non-CAD subjects. For each subject, five-channel heart sound signals were recorded synchronously for 5 min. After data segmentation and quality evaluation, 553 samples were left in the CAD group and 438 samples in the non-CAD group. The time-domain, frequency-domain, entropy, and cross entropy features were extracted. After feature selection, the optimal feature set was fed into the support vector machine for classification. The results showed that from single-channel to multi-channel, the classification accuracy has increased from 78.75% to 86.70%. After adding entropy features and cross entropy features, the classification accuracy continued to increase to 90.92%. The study indicated that the method based on multi-domain feature fusion of multi-channel heart sound signals could provide more information for CAD detection, and entropy features and cross entropy features played an important role in it.

Continuous and Noninvasive Estimation of Right Ventricle Systolic Blood Pressure Using Heart Sound Signal by Deep Bidirectional LSTM Network

Objective: Timely monitoring right ventricular systolic blood pressure (RVSBP) is helpful in the early detection of pulmonary hypertension (PH). However, it is not easy to monitor RVSBP directly. The objective of this paper is to develop a deep learning technique for RVSBP noninvasive estimation using heart sound (HS) signals supported by (electrocardiography) ECG signals without complex features extraction. Methods: Five beagle dog subjects were used. The medicine U-44069 was injected into the subjects to induce a wide range of RVSBP variation. The blood pressure in right ventricle, ECG of lead I and HS signals were recorded simultaneously. Thirty-two records were collected. The relations between RVSBP and cyclic HS signals were modeled by the Bidirectional Long Short-Term Memory (Bi-LSTM) network. Results: The mean absolute error (MAE) ± standard deviation (SD) inside record was 1.85 ± 1.82 mmHg. It was 4.37 ± 2.49 mmHg across record but within subject. The corrective factors were added after training the Bi-LSTM network across subjects. Finally, the MAE ± SD from 12.46 ± 6.56 mmHg dropped to 6.37 ± 4.90 mmHg across subjects. Significance: Our work was the first to apply the Bi-LSTM network to build relations between the HS signal and RVSBP. This work suggested a noninvasive and continuous RVSBP estimation using the HS signal supported by the ECG signal by deep learning architecture without the need of healthcare professionals.

Entropy Measures in Analysis of Head up Tilt Test Outcome for Diagnosing Vasovagal Syncope

The paper presents possible applications of entropy measures in analysis of biosignals recorded during head up tilt testing (HUTT) in patients with suspected vasovagal syndrome. The study group comprised 80 patients who developed syncope during HUTT (57 in the passive phase of the test (HUTT(+) group) and 23 who had negative result of passive phase and developed syncope after provocation with nitroglycerine (HUTT(−) group)). The paper focuses on assessment of monitored signals’ complexity (heart rate expressed as R-R intervals (RRI), blood pressure (sBP, dBP) and stroke volume (SV)) using various types of entropy measures (Sample Entropy (SE), Fuzzy Entropy (FE), Shannon Entropy (Sh), Conditional Entropy (CE), Permutation Entropy (PE)). Assessment of the complexity of signals in supine position indicated presence of significant differences between HUTT(+) versus HUTT(−) patients only for Conditional Entropy (CE(RRI)). Values of CE(RRI) higher than 0.7 indicate likelihood of a positive result of HUTT already at the passive phase. During tilting, in the pre-syncope phase, significant differences were found for: (SE(sBP), SE(dBP), FE(RRI), FE(sBP), FE(dBP), FE(SV), Sh(sBP), Sh(SV), CE(sBP), CE(dBP)). HUTT(+) patients demonstrated significant changes in signals’ complexity more frequently than HUTT(−) patients. When comparing entropy measurements done in the supine position with those during tilting, SV assessed in HUTT(+) patients was the only parameter for which all tested measures of entropy (SE(SV), FE(SV), Sh(SV), CE(SV), PE(SV)) showed significant differences.