Asunto(s)

RESUMEN

There has been a proliferation of machine learning (ML) electrocardiogram (ECG) classification algorithms reaching >85% accuracy for various cardiac pathologies. Despite the high accuracy at individual institutions, challenges remain when it comes to multi-center deployment. Transfer learning (TL) is a technique in which a model trained for a specific task is repurposed for another related task, in this case ECG ML model trained at one institution is fine-tuned to be utilized to classify ECGs at another institution. Models trained at one institution, however, might not be generalizable for accurate classification when deployed broadly due to differences in type, time, and sampling rate of traditional ECG acquisition. In this study, we evaluate the performance of time domain (TD) and frequency domain (FD) convolutional neural network (CNN) classification models in an inter-institutional scenario leveraging three different publicly available datasets. The larger PTB-XL ECG dataset was used to initially train TD and FD CNN models for atrial fibrillation (AFIB) classification. The models were then tested on two different data sets, Lobachevsky University Electrocardiography Database (LUDB) and Korea University Medical Center database (KURIAS). The FD model was able to retain most of its performance (>0.81 F1-score), whereas TD was highly affected (<0.53 F1-score) by the dataset variations, even with TL applied. The FD CNN showed superior robustness to cross-institutional variability and has potential for widespread application with no compromise to ECG classification performance.

Asunto(s)

RESUMEN

There has been a proliferation of machine learning (ML) electrocardiogram (ECG) classification algorithms reaching > 85% accuracy for various cardiac pathologies. Although the accuracy within institutions might be high, models trained at one institution might not be generalizable enough for accurate detection when deployed in other institutions due to differences in type of signal acquisition, sampling frequency, time of acquisition, device noise characteristics and number of leads. In this proof-of-concept study, we leverage the publicly available PTB-XL dataset to investigate the use of time-domain (TD) and frequency-domain (FD) convolutional neural networks (CNN) to detect myocardial infarction (MI), ST/T-wave changes (STTC), atrial fibrillation (AFIB) and sinus arrhythmia (SARRH). To simulate interinstitutional deployment, the TD and FD implementations were also compared on adapted test sets using different sampling frequencies 50 Hz, 100 Hz and 250 Hz, and acquisition times of 5 s and 10s at 100 Hz sampling frequency from the training dataset. When tested on the original sampling frequency and duration, the FD approach showed comparable results to TD for MI (0.92 FD - 0.93 TD AUROC) and STTC (0.94 FD - 0.95 TD AUROC), and better performance for AFIB (0.99 FD - 0.86 TD AUROC) and SARRH (0.91 FD - 0.65 TD AUROC). Although both methods were robust to changes in sampling frequency, changes in acquisition time were detrimental to the TD MI and STTC AUROCs, at 0.72 and 0.58 respectively. Alternatively, the FD approach was able to maintain the same level of performance, and, therefore, showed better potential for interinstitutional deployment.

Asunto(s)

RESUMEN

Pancreatic cancer is the fourth most common cause of cancer-related fatalities as there are a limited number of tools to diagnose this disease in its early stages. Pancreatitis is characterized as an inflammation of the pancreatic tissue due to an excess amount of pancreatic enzymes remaining in the organ. Both of these diseases result in a stiffening of the tissue which makes them suitable for the use of elastography techniques as a diagnostic method. However, these methods typically assume that the tissue is purely elastic when biological tissue is inherently viscoelastic. The attenuation measuring ultrasound shear elastography (AMUSE) method, which measures both attenuation and shear wave velocity was used to characterize the viscoelasticity of pancreatic tissue. This method was tested in ex vivo normal porcine samples that were also stiffened in formalin and in vivo by conducting studies in healthy human subjects. Ex vivo testing showed ranges of phase velocity, group velocity, and phase attenuation values of 1.05 - 1.33 m/s, 0.83 - 1.12 m/s, and 183 - 210 Np/m. After immersing the ex vivo tissue in formalin there was a distinguishable difference between normal and stiffened tissue. This study produced percent difference ranges of phase velocity, group velocity, and phase attenuation from 0 to 100 minutes in formalin of 30.0% - 56.5%, 38.2% - 58.6%, and 55.8% - 64.8%, respectively. The ranges of phase velocity, group velocity, and phase attenuation results in human subjects were 1.53 - 1.60 m/s, 1.76 - 1.91 m/s, and 196 - 204 Np/m, respectively. These results were within a similar range reported by other elastography techniques. Further work with the AMUSE method in subjects with pancreatitis and cancer is needed to determine its effectiveness in showing a difference between healthy and diseased tissue in humans.

RESUMEN

Diastolic dysfunction causes close to half of congestive heart failures and is associated with increased stiffness in left-ventricular myocardium. A clinical tool capable of measuring viscoelasticity of the myocardium could be beneficial in clinical settings. We used Lamb wave Dispersion Ultrasound Vibrometry (LDUV) for assessing the feasibility of making in vivo non-invasive measurements of myocardial elasticity and viscosity in pigs. In vivo open-chest measurements of myocardial elasticity and viscosity obtained using a Fourier space based analysis of Lamb wave dispersion are reported. The approach was used to perform ECG-gated transthoracic in vivo measurements of group velocity, elasticity and viscosity throughout a single heart cycle. Group velocity, elasticity and viscosity in the frequency range 50-500 Hz increased from diastole to systole, consistent with contraction and relaxation of the myocardium. Systolic group velocity, elasticity and viscosity were 5.0 m/s, 19.1 kPa, 6.8 Pa·s, respectively. In diastole, the measured group velocity, elasticity and viscosity were 1.5 m/s, 5.1 kPa and 3.2 Pa·s, respectively.

RESUMEN

Patients with congenital heart disease with a pressure-overloaded right ventricle can develop liver disease and would benefit from non-invasive diagnostic modalities such as ultrasound shear wave elastography (US SWE). We sought to investigate the ability of US SWE to measure dynamic changes in liver stiffness with an acute fluid bolus in an animal model. Three piglets underwent surgical intervention to create a pressure-overloaded right ventricle and, 12 wk later, underwent US SWE, both pre- and post-intravenous infusion of a saline bolus. Ultrasound measures of shear modulus, velocity and attenuation were taken to characterize hepatic mechanical properties. Liver stiffness exhibited a dynamic component that increased after fluid bolus, although not reaching statistical significance with our small sample size, and these changes were greater in more diseased livers. US SWE may provide a promising non-invasive method for assessing dynamic changes in hydration status and degree of liver disease.

Asunto(s)

RESUMEN

Currently, dynamic elastography techniques estimate the linear elastic shear modulus of different body tissues. New methods that investigate other properties of soft tissues such as anisotropy, viscosity, and shear nonlinearity would provide more information about the structure and function of the tissue and might provide a better contrast than tissue stiffness and hence provide more effective diagnostic tools for some diseases. It has previously been shown that shear wave velocity in a medium changes due to an applied stress, a phenomenon called acoustoelasticity (AE). Applying a stress to compress a medium while measuring the shear wave velocity versus strain provides data with which the third-order nonlinear shear modulus can be estimated. To evaluate the feasibility of estimating , we evaluated ten ex vivo porcine kidneys embedded in 10% porcine gelatin to mimic the case of a transplanted kidney. Under assumptions of an elastic incompressible medium for AE measurements, the shear modulus was quantified at each compression level and the applied strain was assessed by measuring the change in the thickness of the kidney cortex. Finally, was calculated by applying the AE theory. Our results demonstrated that it is possible to estimate a nonlinear shear modulus by monitoring the changes in strain and due to kidney deformation. The magnitudes of are higher when the compression is performed progressively and when using a plate attached to the transducer. Nevertheless, the values obtained for are similar to those previously reported in the literature for breast tissue.

Asunto(s)

RESUMEN

Ultrasound and magnetic resonance elastography techniques are used to assess mechanical properties of soft tissues. Tissue stiffness is related to various pathologies such as fibrosis, loss of compliance, and cancer. One way to perform elastography is measuring shear wave velocity of propagating waves in tissue induced by intrinsic motion or an external source of vibration, and relating the shear wave velocity to tissue elasticity. All tissues are inherently viscoelastic and ignoring viscosity biases the velocity-based estimates of elasticity and ignores a potentially important parameter of tissue health. We present attenuation measuring ultrasound shearwave elastography (AMUSE), a technique that independently measures both shear wave velocity and attenuation in tissue and therefore allows characterization of viscoelasticity without using a rheological model. The theoretical basis for AMUSE is first derived and validated in finite element simulations. AMUSE is validated against the traditional methods for assessing shear wave velocity (phase gradient) and attenuation (amplitude decay) in tissue mimicking phantoms and excised tissue. The results agreed within one standard deviation. AMUSE was used to measure shear wave velocity and attenuation in 15 transplanted livers in patients with potential acute rejection, and the results were compared with the biopsy findings in a preliminary study. The comparison showed excellent agreement and suggests that AMUSE can be used to separate transplanted livers with acute rejection from livers with no rejection.

Asunto(s)

RESUMEN

PURPOSE: We propose a novel method to monitor bladder wall mechanical properties as a function of filling volume, with the potential application to bladder compliance assessment. The proposed ultrasound bladder vibrometry (UBV) method uses ultrasound to excite and track Lamb waves on the bladder wall from which its mechanical properties are derived by fitting measurements to an analytical model. Of particular interest is the shear modulus of bladder wall at different volumes, which we hypothesize, is similar to measuring the compliance characteristics of the bladder. MATERIALS AND METHODS: Three experimental models were used: 1) an ex vivo porcine model where normal and aberrant (stiffened by formalin) bladders underwent evaluation by UBV; 2) an in vivo study to evaluate the performance of UBV on patients with clinically documented compliant and noncompliant bladders undergoing UDS; and 3) a noninvasive UBV protocol to assess bladder compliance using oral hydration and fractionated voiding on three healthy volunteers. RESULTS: The ex vivo studies showed a high correlation between the UBV parameters and direct pressure measurement (R2 = 0.84-0.99). A similar correlation was observed for 2 patients with compliant and noncompliant bladders (R2 = 0.89-0.99) undergoing UDS detrusor pressure-volume measurements. The results of UBV on healthy volunteers, performed without catheterization, were comparable to a compliant bladder patient. CONCLUSION: The utility of UBV as a method to monitor changes in bladder wall mechanical properties is validated by the high correlation with pressure measurements in ex vivo and in vivo patient studies. High correlation UBV and UDS in vivo studies demonstrated the potential of UBV as a bladder compliance assessment tool. Results of studies on healthy volunteers with normal bladders demonstrated that UBV could be performed noninvasively. Further studies on a larger cohort are needed to fully validate the use of UBV as a clinical tool for bladder compliance assessment.

Asunto(s)

RESUMEN

Esophageal cancer is a malignant neoplasm with poor outcomes. Determination of local disease progression is a major determining factor in treatment modality, radiation dose, radiation field and subsequent surgical therapy. Discrimination of true tumor extent is difficult given the similarity of soft tissues of the malignancy compared to non-malignant tissues using current imaging modalities. A possible method to discriminate between these tissues may be to exploit mechanical properties to diagnostic advantage, as malignant tissues tend to be stiffer relative to normal adjacent tissue. Shear waves propagate faster in stiffer tissues relative to softer tissues. This may be measured by using ultrasound based shear wave vibrometry. In this method, acoustic radiation force is used to create a shear wave in the tissue of interest and ultrafast ultrasound imaging is used to track the propagating wave to measure the wave velocity and estimate the shear moduli. In this study we created simulated malignant lesions (1.5 cm length) using radiofrequency ablation in ex vivo esophageal samples with varied progression (partial thickness n = 4, and full thickness n = 5) and used normal regions of the same esophageal specimen as controls. Shear wave vibrometry was used to measure shear wave group velocity and shear wave phase velocity in the ex vivo specimens. These values were used to estimate shear moduli using an elastic shear wave model and elastic and viscoelastic Lamb wave models. Our results show that the group and phase velocities increase due to both full and mucosal ablation, and that discrimination may be provided by higher order analysis using viscoelastic Lamb wave fitting. This technique may have application for determination of extent of early esophageal malignancy and warrants further investigation using in vivo approaches to determine performance compared to current imaging modalities.

RESUMEN

The myocardium is known to be an anisotropic medium where the muscle fiber orientation changes through the thickness of the wall. Shear wave elastography methods use propagating waves which are measured by ultrasound or magnetic resonance imaging (MRI) techniques to characterize the mechanical properties of various tissues. Ultrasound- or MR-based methods have been used and the excitation frequency ranges for these various methods cover a large range from 24-500 Hz. Some of the ultrasound-based methods have been shown to be able to estimate the fiber direction. We constructed a model with layers of elastic, transversely isotropic materials that were oriented at different angles to simulate the heart wall in systole and diastole. We investigated the effect of frequency on the wave propagation and the estimation of fiber direction and wave speeds in the different layers of the assembled models. We found that waves propagating at low frequencies such as 30 or 50 Hz showed low sensitivity to the fiber direction but also had substantial bias in estimating the wave speeds in the layers. Using waves with higher frequency content (>200 Hz) allowed for more accurate fiber direction and wave speed estimation. These results have particular relevance for MR- and ultrasound-based elastography applications in the heart.

Asunto(s)

RESUMEN

Evaluation of tissue engineering constructs is performed by a series of different tests. In many cases it is important to match the mechanical properties of these constructs to those of native tissues. However, many mechanical testing methods are destructive in nature which increases cost for evaluation because of the need for additional samples reserved for these assessments. A wave propagation method is proposed for characterizing the shear elasticity of thin layers bounded by a rigid substrate and fluid-loading, similar to the configuration for many tissue engineering applications. An analytic wave propagation model was derived for this configuration and compared against finite element model simulations and numerical solutions from the software package Disperse. The results from the different models found very good agreement. Experiments were performed in tissue-mimicking gelatin phantoms with thicknesses of 1 and 4 mm and found that the wave propagation method could resolve the shear modulus with very good accuracy, no more than 4.10% error. This method could be used in tissue engineering applications to monitor tissue engineering construct maturation with a nondestructive wave propagation method to evaluate the shear modulus of a material.

Asunto(s)

RESUMEN

Tissues such as skeletal muscle and kidneys have well-defined structure that affects the measurements of mechanical properties. As an approach to characterize the material properties of these tissues, different groups have assumed that they are transversely isotropic (TI) and measure the shear wave velocity as it varies with angle with respect to the structural architecture of the organ. To refine measurements in these organs, it is desirable to have tissue-mimicking phantoms that exhibit similar anisotropic characteristics. Some approaches involve embedding fibers into a material matrix. However, if a homogeneous solid is under compression due to a static stress, an acoustoelastic effect can manifest that makes the measured wave velocities change with the compression stress. We propose to exploit this characteristic to demonstrate that stressed tissue mimicking phantoms can be characterized as a TI material. We tested six phantoms made with different concentrations of gelatin and agar. Stress was applied by the weight of a water container centered on top of a plate on top of the phantom. A linear array transducer and a V-1 Verasonics system were used to induce and measure shear waves in the phantoms. The shear wave motion was measured using a compound plane wave imaging technique. Autocorrelation was applied to the received in-phase/quadrature data. The shear wave velocity, c, was estimated using a Radon transform method. The transducer was mounted on a rotating stage so measurements were made every 10° over a range of 0° to 360°, where the stress is applied along 0° to 180° direction. The shear moduli were estimated. A TI model was fit to the data and the fractional anisotropy was evaluated. This approach can be used to explore many configurations of transverse isotropy with the same phantom, simply by applying stress to the tissue-mimicking phantom.

Asunto(s)

RESUMEN

Ultrasound radiation force-based methods can quantitatively evaluate tissue viscoelastic material properties. One of the limitations of the current methods is neglecting the inherent anisotropy nature of certain tissues. To explore the phenomenon of anisotropy in a laboratory setting, we created two phantom designs incorporating fibrous and fishing line material with preferential orientations. Four phantoms were made in a cube-shaped mold; both designs were arranged in multiple layers and embedded in porcine gelatin using two different concentrations (8%, 14%). An excised sample of pork tenderloin was also studied. Measurements were made in the phantoms and the pork muscle at different angles by rotating the phantom with respect to the transducer, where 0° and 180° were defined along the fibers, and 90° and 270° across the fibers. Shear waves were generated and measured by a Verasonics ultrasound system equipped with a linear array transducer. For the fibrous phantom, the mean and standard deviations of the shear wave speeds along (0°) and across the fibers (90°) with 8% gelatin were 3.60  ±  0.03 and 3.18  ±  0.12 m s(-1) and with 14% gelatin were 4.10  ±  0.11 and 3.90  ±  0.02 m s(-1). For the fishing line material phantom, the mean and standard deviations of the shear wave speeds along (0°) and across the fibers (90°) with 8% gelatin were 2.86  ±  0.20 and 2.44  ±  0.24 m s(-1) and with 14% gelatin were 3.40  ±  0.09 and 2.84  ±  0.14 m s(-1). For the pork muscle, the mean and standard deviations of the shear wave speeds along the fibers (0°) at two different locations were 3.83  ±  0.16 and 3.86  ±  0.12 m s(-1) and across the fibers (90°) were 2.73  ±  0.18 and 2.70  ±  0.16 m s(-1), respectively. The fibrous and fishing line gelatin-based phantoms exhibited anisotropy that resembles that observed in the pork muscle.

Asunto(s)

RESUMEN

Catheters are increasingly used therapeutically and investigatively. With complex usage comes a need for more accurate intracardiac localization than traditional guidance can provide. An injection catheter navigated by ultrasound was designed and then tested in an open-chest model of acute ischemia in eight pigs. The catheter is made "acoustically active" by a piezo-electric crystal near its tip, electronically controlled, vibrating in the audio frequency range and uniquely identifiable using pulsed-wave Doppler. Another "target" crystal was sutured to the epicardium within the ischemic region. Sonomicrometry was used to measure distances between the two crystals and then compared with measurements from 2-D echocardiographic images. Complete data were obtained from seven pigs, and the correlation between sonomicrometry and ultrasound measurements was excellent (p < 0.0001, ρ = 0.9820), as was the intraclass correlation coefficient (0.96) between two observers. These initial experimental results suggest high accuracy of ultrasound navigation of the acoustically active catheter prototype located inside the beating left ventricle.

Asunto(s)

RESUMEN

A healthy compliant bladder is capable of storing increasing volumes of urine at low pressures. The loss of bladder compliance is associated with various diseases. The urodynamic studies (UDS), the current clinical gold standard for measuring bladder compliance, requires catheterization and measuring intra-bladder pressure as a function of filling volumes. Ultrasound Bladder Vibrometry (UBV) is a noninvasive technique that uses focused ultrasound radiation force to excite Lamb waves in the bladder wall and pulse-echo techniques to track the wave motion in tissue. Cross-spectral analysis is used to calculate the wave velocity, which is directly related to the elastic properties of the bladder wall. In this study, we compare the measurements of changes in bladder elasticity as a function of bladder pressure and volume obtained using UBV and the pressure-volume measurements obtained using UDS. UBV and UDS of an excised porcine bladder are presented. Comparative studies in neurogenic and healthy patient bladders are also summarized.

Asunto(s)

RESUMEN

Increase in bladder stiffness could be associated with various pathophysiologic conditions. Measuring bladder viscoelasticity could be an important step towards understanding various disease processes and improving patient care. Here, we introduce ultrasound bladder vibrometry (UBV), a novel method for rapid and noninvasive measurement of bladder wall viscoelasticity. UBV uses acoustic radiation force to excite mechanical waves in the bladder wall and track the motion using ultrasound pulse-echo techniques. Fourier domain analysis of the tissue motion versus time is used to calculate the phase velocity dispersion (change of phase velocity as a function of frequency). The measured phase velocity dispersion is fit with the antisymmetric Lamb wave model to estimate tissue elasticity and viscosity. We used finite element analysis of viscoelastic plate deformation to investigate the effect of curvature on Lamb wave dispersion and showed that the effects of curvature are negligible. The feasibility of the UBV technique was demonstrated in ex vivo and in vivo settings. Elasticity and viscosity of excised pig at various filling volumes (V) and pressures (p) were found to be µ1 = 9.6 kPa and µ2 = 0.2 Pa s (V = 187 ml and p = 8.6 mmHg), µ1 = 48.7 kPa and µ2 = 3.5 Pa s (V = 267 ml and p = 17.6 mmHg), and µ1 = 106.9 kPa and µ2 = 1.5 Pa s (V = 327 ml and p = 27.6 mmHg) respectively. Transabdominal measurements in an anesthetized pig found values of bladder elasticity µ1 = 26.1 kPa and viscosity µ2 = 0.9 Pa s and demonstrate the ability of UBV to perform in vivo measurements. The results presented in this paper introduce a novel technique for measuring mechanical properties of the bladder and lay the foundation for further investigation of the effects of pathology on bladder viscoelasticity.

Asunto(s)

Asunto(s)

RESUMEN

Viscoelastic properties of the myocardium are important for normal cardiac function and may be altered by disease. Thus, quantification of these properties may aid with evaluation of the health of the heart. Lamb wave dispersion ultrasound vibrometry (LDUV) is a shear wave-based method that uses wave velocity dispersion to measure the underlying viscoelastic material properties of soft tissue with plate-like geometries. We tested this method in eight pigs in an open-chest preparation. A mechanical actuator was used to create harmonic, propagating mechanical waves in the myocardial wall. The motion was tracked using a high frame rate acquisition sequence, typically 2500 Hz. The velocities of wave propagation were measured over the 50-400 Hz frequency range in 50 Hz increments. Data were acquired over several cardiac cycles. Dispersion curves were fit with a viscoelastic, anti-symmetric Lamb wave model to obtain estimates of the shear elasticity, µ(1), and viscosity, µ(2) as defined by the Kelvin-Voigt rheological model. The sensitivity of the Lamb wave model was also studied using simulated data. We demonstrated that wave velocity measurements and Lamb wave theory allow one to estimate the variation of viscoelastic moduli of the myocardial walls in vivo throughout the course of the cardiac cycle.

Asunto(s)

RESUMEN

Characterization of the viscoelastic material properties of soft tissue has become an important area of research over the last two decades. Our group has been investigating the feasibility of using a shear wave dispersion ultrasound vibrometry (SDUV) method to excite Lamb waves in organs with plate-like geometry to estimate the viscoelasticity of the medium of interest. The use of Lamb wave dispersion ultrasound vibrometry to quantify the mechanical properties of viscoelastic solids has previously been reported. Two organs, the heart wall and the spleen, can be readily modeled using plate-like geometries. The elasticity of these two organs is important because they change in pathological conditions. Diastolic dysfunction is the inability of the left ventricle (LV) of the heart to supply sufficient stroke volumes into the systemic circulation and is accompanied by the loss of compliance and stiffening of the LV myocardium. It has been shown that there is a correlation between high splenic stiffness in patients with chronic liver disease and strong correlation between spleen and liver stiffness. Here, we investigate the use of the SDUV method to quantify the viscoelasticity of the LV free-wall myocardium and spleen by exciting Rayleigh waves on the organ's surface and measuring the wave dispersion (change of wave velocity as a function of frequency) in the frequency range 40­500 Hz. An equation for Rayleigh wave dispersion due to cylindrical excitation was derived by modeling the excised myocardium and spleen with a homogenous Voigt material plate immersed in a nonviscous fluid. Boundary conditions and wave potential functions were solved for the surface wave velocity. Analytical and experimental convergence between the Lamb and Rayleigh waves is reported in a finite element model of a plate in a fluid of similar density, gelatin plate and excised porcine spleen and left-ventricular free-wall myocardium.

Asunto(s)