Application of adaptive filters to noninvasive acoustical detection of coronary occlusions before and after angioplasty

Custom-designed sensors embedded 3D-printed wearable device for improving the hemodialysis-related vascular dysfunction detection

BACKGROUND

The increasing prevalence of end-stage renal disease (ESRD) imposes a substantial economic burden on public health-care systems. Hemodialysis (HD) is a pivotal treatment modality for patients with ESRD. However, prolonged use of HD vessels may result in stenosis, thrombosis, and occlusion due to repeated daily punctures. Thus, early detection and prevention of the dysfunction of dialysis routes are crucial.

OBJECTIVE

In this study, we designed a wearable device for the early and accurate detection of arteriovenous access (AVA) stenosis in HD patients.

METHODS

A personalized three-dimensional (3D) printed wearable device was designed by combining the phonoangiography (PAG) and photoplethysmography (PPG) techniques. The capability of this device to monitor AVA dysfunction before and after percutaneous transluminal angioplasty (PTA) was evaluated.

RESULTS

After PTA, the amplitudes of both PAG and PPG signals increased in patients with arteriovenous fistulas and those with arteriovenous grafts; this might be due to increased blood flow.

CONCLUSION

Our designed multi-sensor wearable medical device using PAG, PPG, and 3D printing appears suitable for early and accurate detection of AVA stenosis in HD patients.

Numerical investigation of wall pressure fluctuations downstream of concentric and eccentric blunt stenosis models

Pressure fluctuations that cause acoustic radiation from vessel models with concentric and eccentric blunt stenoses are investigated. Large eddy simulations of non-pulsatile flow condition are performed using OpenFOAM. Calculated amplitude and spatial-spectral distribution of acoustic pressures at the post-stenotic region are compared with previous experimental and theoretical results. It is found that increasing the Reynolds number does not change the location of the maximum root mean square wall pressure, but causes a general increase in the spectrum level, although the change in the shape of the spectrum is not significant. On the contrary, compared to the concentric model at the same Reynolds number, eccentricity leads to an increase both at the distance of the location of the maximum root mean square wall pressure from the stenosis exit and the spectrum level. This effect becomes more distinct when radial eccentricity of the stenosis increases. Both the flow rate and the eccentricity of the stenosis shape are evaluated to be clinically important parameters in diagnosing stenosis.

The clinical evaluation of the CADence device in the acoustic detection of coronary artery disease

The noninvasive detection of turbulent coronary flow may enable diagnosis of significant coronary artery disease (CAD) using novel sensor and analytic technology. Eligible patients (n = 1013) with chest pain and CAD risk factors undergoing nuclear stress testing were studied using the CADence (AUM Cardiovascular Inc., Northfield MN) acoustic detection (AD) system. The trial was designed to demonstrate non-inferiority of AD for diagnostic accuracy in detecting significant CAD as compared to an objective performance criteria (sensitivity 83% and specificity 80%, with 15% non-inferiority margins) for nuclear stress testing. AD analysis was blinded to clinical, core lab-adjudicated angiographic, and nuclear data. The presence of significant CAD was determined by computed tomographic (CCTA) or invasive angiography. A total of 1013 subjects without prior coronary revascularization or Q-wave myocardial infarction were enrolled. Primary analysis was performed on subjects with complete angiographic and AD data (n = 763) including 111 subjects (15%) with severe CAD based on CCTA (n = 34) and invasive angiography (n = 77). The sensitivity and specificity of AD were 78% (p = 0.012 for non-inferiority) and 35% (p < 0.001 for failure to demonstrate non-inferiority), respectively. AD results had a high 91% negative predictive value for the presence of significant CAD. AD testing failed to demonstrate non-inferior diagnostic accuracy as compared to the historical performance of a nuclear stress OPC due to low specificity. AD sensitivity was non-inferior in detecting significant CAD with a high negative predictive value supporting a potential value in excluding CAD.

Model validation for a noninvasive arterial stenosis detection problem

A current thrust in medical research is the development of a non-invasive method for detection, localization, and characterization of an arterial stenosis (a blockage or partial blockage in an artery). A method has been proposed to detect shear waves in the chest cavity which have been generated by disturbances in the blood flow resulting from a stenosis. In order to develop this methodology further, we use one-dimensional shear wave experimental data from novel acoustic phantoms to validate a corresponding viscoelastic mathematical model. We estimate model parameters which give a good fit (in a sense to be precisely defined) to the experimental data, and use asymptotic error theory to provide confidence intervals for parameter estimates. Finally, since a robust error model is necessary for accurate parameter estimates and confidence analysis, we include a comparison of absolute and relative models for measurement error.

Path length entropy analysis of diastolic heart sounds

Early detection of coronary artery disease (CAD) using the acoustic approach, a noninvasive and cost-effective method, would greatly improve the outcome of CAD patients. To detect CAD, we analyze diastolic sounds for possible CAD murmurs. We observed diastolic sounds to exhibit 1/f structure and developed a new method, path length entropy (PLE) and a scaled version (SPLE), to characterize this structure to improve CAD detection. We compare SPLE results to Hurst exponent, Sample entropy and Multiscale entropy for distinguishing between normal and CAD patients. SPLE achieved a sensitivity-specificity of 80%-81%, the best of the tested methods. However, PLE and SPLE are not sufficient to prove nonlinearity, and evaluation using surrogate data suggests that our cardiovascular sound recordings do not contain significant nonlinear properties.

Acoustical detection of venous stenosis in hemodialysis patients using principal component analysis

In this paper, a feature extraction method based on principal component analysis was developed for classification of the vascular access's condition in hemodialysis patients. The assessment of the method was carried out by discriminating between before and after angioplasty sound recordings as well as before angioplasty and reference recordings. The results showed that when before and after angioplasty recordings were compared by patient, the classification agreed with the result of angioplasty procedure. When all the available before and after angioplasty recordings were compared, it was still possible to discriminate them at a good rate. On the other hand, when the reference recordings substituted the after angioplasty recordings, almost a perfect discrimination was achieved.

Characterisation of arteriovenous fistula's sound recordings using principal component analysis

In this study, a signal analysis framework based on the Karhunen-Loève expansion and k-means clustering algorithm is proposed for the characterisation of arteriovenous (AV) fistula's sound recordings. The Karhunen-Loève (KL) coefficients corresponding to the directions of maximum variance were used as classification features, which were clustered applying k-means algorithm. The results showed that one natural cluster was found for similar AV fistula's state recordings. On the other hand, when stenotic and non-stenotic AV fistula's recordings were processed together, the two most significant KL coefficients contain important information that can be used for classification or discrimination between these AV fistula's states.

Experimental and Computational Models for Simulating Sound Propagation Within the Lungs

An acoustic boundary element model is used to simulate sound propagation in the lung parenchyma and surrounding chest wall. It is validated theoretically and numerically and then compared with experimental studies on lung-chest phantom models that simulate the lung pathology of pneumothorax. Studies quantify the effect of the simulated lung pathology on the resulting acoustic field measured at the phantom chest surface. This work is relevant to the development of advanced auscultatory techniques for lung, vascular, and cardiac sounds within the torso that utilize multiple noninvasive sensors to create acoustic images of the sound generation and transmission to identify certain pathologies.

Acoustic detection of coronary artery disease

Coronary artery disease (CAD) occurs when the arteries to the heart (the coronary arteries) become blocked by deposition of plaque, depriving the heart of oxygen-bearing blood. This disease is arguably the most important fatal disease in industrialized countries, causing one-third to one-half of all deaths in persons between the ages of 35 and 64 in the United States. Despite the fact that early detection of CAD allows for successful and cost-effective treatment of the disease, only 20% of CAD cases are diagnosed prior to a heart attack. The development of a definitive, noninvasive test for detection of coronary blockages is one of the holy grails of diagnostic cardiology. One promising approach to detecting coronary blockages noninvasively is based on identifying acoustic signatures generated by turbulent blood flow through partially occluded coronary arteries. In fact, no other approach to the detection of CAD promises to be as inexpensive, simple to perform, and risk free as the acoustic-based approach. Although sounds associated with partially blocked arteries are easy to identify in more superficial vessels such as the carotids, sounds from coronary arteries are very faint and surrounded by noise such as the very loud valve sounds. To detect these very weak signals requires sophisticated signal processing techniques. This review describes the work that has been done in this area since the 1980s and discusses future directions that may fulfill the promise of the acoustic approach to detecting coronary artery disease.

Contact-Free Measurement of Cardiac Pulse Based on the Analysis of Thermal Imagery

We have developed a novel method to measure human cardiac pulse at a distance. It is based on the information contained in the thermal signal emitted from major superficial vessels. This signal is acquired through a highly sensitive thermal imaging system. Temperature on the vessel is modulated by pulsative blood flow. To compute the frequency of modulation (pulse), we extract a line-based region along the vessel. Then, we apply fast Fourier transform (FFT) to individual points along this line of interest to capitalize on the pulse's thermal propagation effect. Finally, we use an adaptive estimation function on the average FFT outcome to quantify the pulse. We have carried out experiments on a data set of 34 subjects and compared the pulse computed from our thermal signal analysis method to concomitant ground-truth measurements obtained through a standard contact sensor (piezo-electric transducer). The performance of the new method ranges from 88.52% to 90.33% depending on the clarity of the vessel's thermal imprint. To the best of our knowledge, it is the first time that cardiac pulse has been measured several feet away from a subject with passive means.

Boundary element model for simulating sound propagation and source localization within the lungs

An acoustic boundary element (BE) model is used to simulate sound propagation in the lung parenchyma. It is computationally validated and then compared with experimental studies on lung phantom models. Parametric studies quantify the effect of different model parameters on the resulting acoustic field within the lung phantoms. The BE model is then coupled with a source localization algorithm to predict the position of an acoustic source within the phantom. Experimental studies validate the BE-based source localization algorithm and show that the same algorithm does not perform as well if the BE simulation is replaced with a free field assumption that neglects reflections and standing wave patterns created within the finite-size lung phantom. The BE model and source localization procedure are then applied to actual lung geometry taken from the National Library of Medicine's Visible Human Project. These numerical studies are in agreement with the studies on simpler geometry in that use of a BE model in place of the free field assumption alters the predicted acoustic field and source localization results. This work is relevant to the development of advanced auscultatory techniques that utilize multiple noninvasive sensors to construct acoustic images of sound generation and transmission to identify pathologies.

Acoustic radiation from a fluid-filled, subsurface vascular tube with internal turbulent flow due to a constriction

The vibration of a thin-walled cylindrical, compliant viscoelastic tube with internal turbulent flow due to an axisymmetric constriction is studied theoretically and experimentally. Vibration of the tube is considered with internal fluid coupling only, and with coupling to internal-flowing fluid and external stagnant fluid or external tissue-like viscoelastic material. The theoretical analysis includes the adaptation of a model for turbulence in the internal fluid and its vibratory excitation of and interaction with the tube wall and surrounding viscoelastic medium. Analytical predictions are compared with experimental measurements conducted on a flow model system using laser Doppler vibrometry to measure tube vibration and the vibration of the surrounding viscoelastic medium. Fluid pressure within the tube was measured with miniature hydrophones. Discrepancies between theory and experiment, as well as the coupled nature of the fluid-structure interaction, are described. This study is relevant to and may lead to further insight into the patency and mechanisms of vascular failure, as well as diagnostic techniques utilizing noninvasive acoustic measurements.

Computerised analysis of auscultatory sounds associated with vascular patency of haemodialysis access

Vascular access for renal dialysis is a lifeline for about 120 000 individuals in the United States. Stethoscope auscultation of vascular sounds has some utility in the assessment of access patency, yet can be highly skill-dependent. The objective of the study was to identify acoustic parameters that are related to changes in vascular access patency. The underlying hypothesis is that stenoses of haemodialysis access vessels or grafts cause vascular sound changes that are detectable using computerised data acquisition and analysis. Eleven patients participated in the study. Their vascular sounds were recorded before and after angiography, which was accompanied by angioplasty in most patients. The sounds were acquired using two electronic stethoscopes and then digitised and analysed on a personal computer. Vessel stenosis changes were found to be associated with changes in acoustic amplitude and/or spectral energy distribution. Certain acoustic parameters correlated well (correlation coefficient = 0.98, p < 0.0001) with the change in the degree of stenosis, suggesting that stenosis severity may be predictable from these parameters. Parameters also appeared to be sensitive to modest diameter changes (> 20%), (p < 0.005, Wilcoxon rank sum test). These results suggest that computerised analysis of vascular sounds may be useful in vessel patency surveillance. Further testing using longitudinal studies may be warranted.

Beamformed nearfield imaging of a simulated coronary artery containing a stenosis

This paper is concerned with the potential for the detection and location of an artery containing a partial blockage by exploiting the space-time properties of the shear wave field in the surrounding elastic soft tissue. As a demonstration of feasibility, an array of piezoelectric film vibration sensors is placed on the free surface of a urethane mold that contains a surgical tube. Inside the surgical tube is a nylon constriction that inhibits the water flowing through the tube. A turbulent field develops in and downstream from the blockage, creating a randomly fluctuating pressure on the inner wall of the tube. This force produces shear and compressional wave energy in the urethane. After the array is used to sample the dominant shear wave space-time energy field at low frequencies, a nearfield (i.e., focused) beamforming process then images the energy distribution in the three-dimensional solid. Experiments and numerical simulations are included to demonstrate the potential of this noninvasive procedure for the early identification of vascular blockages-the typical precursor of serious arterial disease in the human heart.

Heart rate variability spectral indices for haemodynamic classification of haemodialysis patients

The usefulness of spectral indices extracted from the heart rate variability (HRV) in discriminating between hypotension-prone and hypotension-resistant haemodialysis patients was investigated. In 30 patients, classified as hypotension resistant (stable group) or hypotension prone (unstable group), beat-to-beat heart period was measured during haemodialysis sessions terminated without collapses. HRV was analysed in the frequency domain combining classic autoregressive spectral estimation with two eigen decomposition-based techniques: the reduced rank approximation (RRA) of the autocorrelation matrix and the Pisarenko harmonic decomposition (PHD). Five spectral indices were obtained: the ratio between the powers in the LF and HF bands (LF/HF), the same ratio calculated after application of RRA (LF/HFRRA), the frequency of the main oscillatory component of HRV estimated through PHD with a decomposition order equal to 1 (F1) and equal to 2 (F2) and the difference between the frequencies of the two oscillatory components resolved in the latter cas (Fd). The performances of these indices in discriminating between the two groups of patients were evaluated estimating the misclassification probability (Pm) of a Bayesian quadratic classifier. The HRV spectral pattern was markedly different: in the stable patients power was mainly in the low-frequency band, whereas in the unstable group it was mainly in the high-frequency band. The frequency of the main oscillatory component was significantly greater in the unstable group than in the stable one. Spectral indices displayed good discrimination power, increasing with the length of the dialysis interval. Best performances were achieved by LF/HFRRA both over short dialysis periods (Pm approximately 12% over 20 min intervals) and over longer periods (Pm = 3.3% over 160 min); similar results were obtained with Fd over short periods and LF/HF over long periods. Spectral HRV indices demonstrate, therefore, a diagnostic value in discriminating between hypotension-resistant and hypotension-prone patients.

Estimation of the ventricular fibrillation duration by autoregressive modeling

An accurate estimation of ventricular fibrillation (VF) duration could be critical in selecting the most effective therapeutic intervention. We test the hypothesis that changes in frequency content of VF signals can be quantified by using autoregressive (AR) modeling, and the duration since the onset of VF can be estimated by using this method. VF signals were recorded for up to 300 s in five isolated rabbit hearts. Fourth-order AR parameters of successive segments were estimated, and frequencies of the first poles (the pole with lower frequency) were pooled together and a curve was fitted: F(t) = A exp (-alpha t) + B, where F(t) is the estimated frequency of the first pole at t'th time instant, alpha is the decay constant, B is the offset frequency, and A is the frequency at time zero minus the offset frequency. The utility of this curve in estimating the VF duration was tested in four new experiments, and the difference between the actual and the estimated VF duration (estimation error) was calculated. F(t), the frequency of the first pole, decreased from 12 to 6 Hz with duration of:VF, while the frequency of the other pole decreased from 25 to 20 Hz. Parameters of the fitted curve were calculated as A = 7.8, alpha = 0.0041 and B was selected as four. Testing on a new set of VF signals produced very little estimation error for the first 100 s of VF, although this error increased with VF duration. For these new signals, the mean value of the absolute estimation error was 26 s. Results of this study show that changes in the frequency content of electrocardiogram (ECG) during VF can be quantified by AR modeling and that the frequency changes associated with a pole of this model can be used to estimate the VF duration.

Model-based analysis of the ECG during early stages of ventricular fibrillation

During the early stages of ventricular fibrillation (VF), identification of the changes in electrocardiographic (ECG) characteristics may be helpful in the determination of defibrillation energy for implantable defibrillators. The hypothesis that the ECG can be quantified by using autoregressive (AR) modeling during VF was tested. Electrocardiograms were recorded for durations of up to 60 seconds of VF in five isolated rabbit hearts. Fourth-order AR parameters of successive 2-second epochs with 50% overlapping of data segments were estimated. At the beginning of VF, mean values of frequency of the first and second poles were 12.5 +/- 1.2 Hz and 24.7 +/- 1.9 Hz, respectively. During VF, frequencies of these poles decreased. At the end of 60 seconds, pole frequencies were 8.7 +/- 1.1 Hz and 21.4 +/- 0.8 Hz. Intersubject variability of the frequencies of the poles was found to be low. Maximum standard deviations for the frequencies of the first and second poles were determined to be 1.9 and 2, respectively. Results of this study show that the VF ECG can be modeled by using the AR modeling technique, and it is possible to quantify the changes in the frequency content of the ECG during VF by using this modeling method.

Mechanoelectrical feedback in cardiac myocytes from stretch-activated ion channels

Stretch-activated ion channels (SAC's) in cardiac myocytes from neonatal rats were studied in cell-attached patches. Stretch of membrane patches by suction in the recording pipette caused the triggering of action potentials that were recorded as action currents (AC's). The significance of a temporal correlation between SAC open probability and AC's was tested using the Kolmogorov-Smirnov and Poisson distributions. It was shown that the 50-ms epoch immediately preceding the action current had unique kinetics and represented a peak in SAC open probability (p < 0.001). Thus it appears that current from a small number of SAC's injects sufficient charge (0.2 pC during 50 ms) to trigger action potentials in myocytes. These data strengthen the hypothesis that passive mechanical stretch of myocardium can be arrhythmogenic.

Noninvasive acoustical detection of coronary artery disease: a comparative study of signal processing methods

Previous studies have indicated heart sounds may contain information useful in the detection of occluded coronary arteries. During diastole, coronary blood flow is maximum, and the sounds associated with turbulent blood flow through partially occluded coronary arteries should be detectable. In order to detect such sounds, recordings of diastolic heart sound segments were analyzed by using four signal processing techniques; the Fast Fourier Transform (FFT), the Autoregressive (AR), the Autoregressive Moving Average (ARMA), and the Minimum-Norm (Eigen-vector) methods. To further enhance the diastolic heart sounds and reduce background noise, an Adaptive filter was used as a preprocessor. The power ratios of the FFT method and the poles of the AR, ARMA, and Eigen-vector methods were used to diagnose patients as diseased or normal arteries using a blind protocol without prior knowledge of the actual disease states of the patients to guard against human bias. Results showed that normal and abnormal records were correctly distinguished in 56 of 80 cases using the Fast Fourier Transform (FFT), in 63 of 80 cases using the AR, in 62 of 80 cases using the ARMA method, and in 67 of 80 cases using the Eigenvector method. Among all four methods, the Eigenvector methods showed the best diagnostic performance when compared with the FFT, AR, and ARMA methods. These results confirm that high frequency acoustic energy between 300 and 800 Hz is associated with coronary stenosis.