Accelerometer Detects Pump Thrombosis and Thromboembolic Events in an In vitro HVAD Circuit

Extracorporeal driveline vibrations to detect left ventricular assist device thrombosis - A porcine model study

BACKGROUND

Pump thrombosis (PT) and related adverse complications contributed to the HeartWare Ventricular Assist Device (HVAD) market withdrawal. Many patients still receive lifelong support, with deficient PT surveillance based on pump power trends. Analysis of pump vibrations is better for detecting PT. Here, we investigated the feasibility of an extracorporeal accelerometer to detect PT from pump vibrations propagated out on the driveline.

METHODS

In a porcine HVAD model (n = 6), an accelerometer was attached to the pump as a reference and another to the driveline for comparisons of signals. In total, 59 thrombi were injected into the heart to induce PT, followed by intermittent thrombus washout maneuvers. Signals were compared visually in spectrograms and quantitatively in third harmonic saliences (S3H) by correlation analysis. Receiver operating characteristic curves expressed the method's outcome in sensitivity vs specificity, with the overall diagnostic performance in the area under the curve (AUC) score.

RESULTS

Five experiments had good driveline signal strength, with clear spectrographic relationships between the 2 accelerometers. Third harmonic driveline vibrations were visible 20 vs 30 times in the reference. The comparison in S3H showed a strong correlation and yielded an AUC of 0.85. Notably, S3H proved robust regarding noise and false PT detections.

CONCLUSIONS

An extracorporeal accelerometer on the driveline can be a readily available method for accurate HVAD PT detection before an accelerometer integration with left ventricular assist device is feasible.

Vibrational characterization of cavitation in left ventricular assist device

BACKGROUND

The left ventricular assist device (LVAD) is a mechanical circulatory support device for patients with severe heart failure. Microbubbles caused by cavitation in the LVAD can potentially lead to physiological and pump-related complications. The aim of this study is to characterize the vibrational patterns in the LVAD during cavitation.

METHODS

The LVAD was integrated into an in vitro circuit and mounted with a high-frequency accelerometer. Accelerometry signals were acquired with different relative pump inlet pressures ranging from baseline (+20 mmHg) to -600 mmHg in order to induce cavitation. Microbubbles were monitored with dedicated sensors at the pump inlet and outlet to quantify the degree of cavitation. Acceleration signals were analyzed in the frequency domain to identify changes in the frequency patterns when cavitation occurred.

RESULTS

Significant cavitation occurred at the low inlet pressure (-600 mmHg) and was detected in the frequency range between 1800 and 9000 Hz. Minor degrees of cavitation at higher inlet pressures (-300 to -500 mmHg) were detected in the frequency range between 500-700, 1600-1700 Hz, and around 12 000 Hz. The signal power of the dominating frequency ranges was statistically significantly different from baseline signals.

CONCLUSION

Vibrational measurements in the LVAD can be used to detect cavitation. A significant degree of cavitation could be detected in a wide frequency range, while minor cavitation activity could only be detected in more narrow frequency ranges. Continuous vibrational LVAD monitoring can potentially be used to detect cavitation and minimize the damaging effect associated with cavitation.

Detection of inflow obstruction in left ventricular assist devices by accelerometer: A porcine model study

BACKGROUND

Left ventricular assist devices (LVAD) provide circulatory blood pump support for severe heart failure patients. Pump inflow obstructions may lead to stroke and pump malfunction. We aimed to verify in vivo that gradual inflow obstructions, representing prepump thrombosis, are detectable by a pump-attached accelerometer, where the routine use of pump power (PLVAD) is deficient.

METHOD

In a porcine model (n = 8), balloon-tipped catheters obstructed HVAD inflow conduits by 34% to 94% in 5 levels. Afterload increases and speed alterations were conducted as controls. We computed nonharmonic amplitudes (NHA) of pump vibrations captured by the accelerometer for the analysis. Changes in NHA and PLVAD were tested by a pairwise nonparametric statistical test. Detection sensitivities and specificities were investigated by receiver operating characteristics with areas under the curves (AUC).

RESULTS

NHA remained marginally affected during control interventions, unlike PLVAD. NHA elevated during obstructions within 52-83%, while mass pendulation was most pronounced. Meanwhile, PLVAD changed far less. Increased pump speeds tended to amplify the NHA elevations. The corresponding AUC was 0.85-1.00 for NHA and 0.35-0.73 for PLVAD.

CONCLUSION

Elevated NHA provides a reliable indication of subclinical gradual inflow obstructions. The accelerometer can potentially supplement PLVAD for earlier warnings and localization of pump.

Prevention of thrombus formation in blood pump by mechanical circular orbital excitation of impeller in magnetically levitated centrifugal pump

BACKGROUND

Mechanical circulatory support devices, such as left ventricular assist devices, have recently been used in patients with heart failure as destination therapy but the formation of thrombus in blood pumps remains a critical problem. In this study, we propose a mechanical antithrombogenic method by impeller excitation using a magnetically levitated (Maglev) centrifugal pump. Previous studies have shown that one-directional excitation prevents thrombus; however, it is effective in only one direction. In this study, we aimed to obtain a better effect by vibrating it in a circular orbit to induce uniform changes in the shear-rate field entirely around the impeller.

METHODS

The blood coagulation time was compared using porcine blood. (1) The flow rate was set to 1 L/min, and applied excitation was at a frequency of 280 Hz and amplitude of 3 μm. (2) Moreover, the effect was compared by varying the frequency, amplitude, and direction of the excitation. In this experiment, the flow rate was set to 0.3 L/min.

RESULTS

(1) The thrombus formation time was 77 min without excitation and 133 min with excitation, which was 1.7 times longer. (2) The results showed no difference between (280 Hz, 3 μm) and (50 Hz, 16 μm) circular orbital excitations, and no directional difference, with thrombus formation of 2.5 times longer under all conditions than that without excitation.

CONCLUSION

In the case of simple reciprocating excitation, the time was approximately 1.2 times longer. This indicated that the circular orbital excitation is more effective.

Detection of inflow obstruction in left ventricular assist devices by accelerometer: An in vitro study

Inflow obstruction in left ventricular assist devices (LVAD) may lead to embolic stroke and pump malfunction. We investigated if an accelerometer detected graded LVAD inflow obstructions. Detection performances were compared to the current continuous surveillance routine based on the pump power consumption (P_LVAD). In ten mock circuit experiments, four different-sized pendulating balloons obstructed HVAD™ inflow conduits cross-section areas by 14%-75%. Nonharmonic amplitudes (NHA) of continuous signals from a triaxial accelerometer attached to the LVAD were compared against single-point P_LVAD values, using load and speed alterations as control interventions. We analyzed the NHA band power with a pairwise nonparametric statistical test. The detection performances were analyzed by receiver operating characteristics with areas under the curves (AUC). The NHA remained unaffected during load alterations. In contrast, NHA increased significantly from the 27% obstruction level (AUC≥0.82), an effect amplified by increased pump speed. P_LVAD did not change significantly below the maximal 75% obstruction level (AUC≤0.36). In conclusion, NHA detected the inflow obstructions much better than P_LVAD. The technique may provide a future monitoring modality of any pendulating obstructive inflow pathology.

Multi-indicator analysis of mechanical blood damage with five clinical ventricular assist devices

PURPOSE

Device-induced blood damage contributes the hemolysis, thrombosis and bleeding complications in patients supported with ventricular assist device (VAD). This study aims to use a multi-indicator method to understand how devices causes blood damage and identify the "hot spots" of blood trauma within VADs.

METHODS

Computational fluid dynamics (CFD) methods were chosen to investigate the hemodynamic features of five clinical VADs (Impella 5.0, UltraMag, CHVAD, HVAD, and HeartMate II) under the same clinical support condition (flow rate of 4.5L/min, pressure head around 75 mmHg). A comprehensive multi-indicator evaluation method including hemodynamic parameters, hemolysis model, thrombotic potential model and bleeding probability model was used to analyze blood damage and assess the hemodynamic performance and hemocompatibility of these VADs.

RESULTS

Simulation results show that shear stress from 50 Pa to 100 Pa plays a major role in blood damage in Impella 5.0, UltraMag and CHVAD, while blood damage in HVAD and HeartMate II is mainly caused by shear stress greater than 100 Pa. Residence time was not the main factor for blood damage in Impella 5.0, and also makes a limited contribution to blood trauma in UltraMag and CHVAD, while it takes a critical role in elevating thrombotic potential in HVAD and HeartMate II. The distribution of regions of high hemolysis risk and high bleeding probability was similar for all these VADs and partially overlapped for high thrombotic potential regions. For Impella 5.0, regions with high hemolysis and bleeding risk were found mainly in the blade tip clearance and diffuser domains, high thrombotic potential regions were almost absent. For UltraMag, regions with high hemolysis, bleeding and thrombosis potential were found in two corners of the inlet pipe, the secondary flow passage, and the impeller eye. For CHVAD, the high-risk regions for hemolysis, bleeding and thrombosis are mainly in the inner side of the secondary flow passage and the middle region of the impeller passage. The narrow hydrodynamic clearance and impeller passage had a high risk of hemolysis and bleeding, and the clearance between the rotor and guide cone and the hydrodynamic clearance had high thrombotic potential. For HeartMate II, regions of high hemolysis risk and bleeding probability were found in the near-wall region of the straightener, the blade tip clearance and the diffuser domain. The corners of the inlet and outlet pipe and the straightener and diffuser regions had high thrombotic potential.

CONCLUSION

The risk of hemolysis, bleeding and thrombosis for these five VADs, in increasing order, was Impella 5.0, UltraMag, CHVAD, HVAD, and HeartMate II. Flow losses caused by the rotor mechanical movement, chaotic flow and narrow clearances increase the blood damage for all these VADs. The multi-indicator analysis can comprehensively evaluate the VAD performance with improved assessment accuracy of CFD.

A New Mathematical Numerical Model to Evaluate the Risk of Thrombosis in Three Clinical Ventricular Assist Devices

(1) Background: Thrombosis is the main complication in patients supported with ventricular assist devices (VAD). Models that accurately predict the risk of thrombus formation in VADs are still lacking. When VADs are clinically assisted, their complex geometric configuration and high rotating speed inevitably generate complex flow fields and high shear stress. These non-physiological factors can damage blood cells and proteins, release coagulant factors and trigger thrombosis. In this study, a more accurate model for thrombus assessment was constructed by integrating parameters such as shear stress, residence time and coagulant factors, so as to accurately assess the probability of thrombosis in three clinical VADs. (2) Methods: A mathematical model was constructed to assess platelet activation and thrombosis within VADs. By solving the transport equation, the influence of various factors such as shear stress, residence time and coagulation factors on platelet activation was considered. The diffusion equation was applied to determine the role of activated platelets and substance deposition on thrombus formation. The momentum equation was introduced to describe the obstruction to blood flow when thrombus is formed, and finally a more comprehensive and accurate model for thrombus assessment in patients with VAD was obtained. Numerical simulations of three clinically VADs (CH-VAD, HVAD and HMII) were performed using this model. The simulation results were compared with experimental data on platelet activation caused by the three VADs. The simulated thrombogenic potential in different regions of MHII was compared with the frequency of thrombosis occurring in the regions in clinic. The regions of high thrombotic risk for HVAD and HMII observed in experiments were compared with the regions predicted by simulation. (3) Results: It was found that the percentage of activated platelets within the VAD obtained by solving the thrombosis model developed in this study was in high agreement with the experimental data (r² = 0.984), the likelihood of thrombosis in the regions of the simulation showed excellent correlation with the clinical statistics (r² = 0.994), and the regions of high thrombotic risk predicted by the simulation were consistent with the experimental results. Further study revealed that the three clinical VADs (CH-VAD, HVAD and HMII) were prone to thrombus formation in the inner side of the secondary flow passage, the clearance between cone and impeller, and the corner region of the inlet pipe, respectively. The risk of platelet activation and thrombus formation for the three VADs was low to high for CH-VAD, HVAD, and HM II, respectively. (4) Conclusions: In this study, a more comprehensive and accurate thrombosis model was constructed by combining parameters such as shear stress, residence time, and coagulation factors. Simulation results of thrombotic risk received with this model showed excellent correlation with experimental and clinical data. It is important for determining the degree of platelet activation in VAD and identifying regions prone to thrombus formation, as well as guiding the optimal design of VAD and clinical treatment.

Acoustic Properties of Axial and Centrifugal Flow Left Ventricular Assist Devices and Prediction of Pump Thrombosis

OBJECTIVE

To characterize the properties of the audible tones produced by current left ventricular assist device (LVAD) pumps approved for use, and to ascertain if changes in those may be present in the setting of pump thrombosis.

PATIENTS AND METHODS

From August 31, 2016, to January 16, 2020, LVAD recipients consented to have surface recordings obtained using a high-fidelity digital stethoscope. Audio data were analyzed using digital recording and editing software to produce an acoustic spectrogram by Fast Fourier transformation.

RESULTS

Recordings were obtained in 53 patient encounters (27 HeartMate II, 19 HeartWare and 7 HeartMate 3). In 12 patients (9 HeartMate II, 3 HeartWare) there was a clinical concern for pump thrombosis. In all patients and pump models, a fundamental frequency was noted, and the second and third harmonics were also clearly detectable. Where thrombosis occurred in the HeartMate II pump, the absolute (normal -46.9 [-57.5,-42.9] dB vs thrombosis -41.4 [-49.8,-26.8] dB; P=.08) and relative (normal 0.72 [0.62, 0.92] vs thrombosis 0.95 [0.86, 1.24]; P=.01) third harmonic frequencies were increased in amplitude. Where paired data were available, an increase in the absolute and relative third harmonic frequencies was observed in all patients. In the case of the HeartWare device, a consistent difference in harmonic amplitudes in the setting of thrombosis could not be identified.

CONCLUSION

A consistent pattern of fundamental and harmonic frequencies is common to all LVADs currently approved for use. Alterations in the amplitude of higher order harmonics may signal the onset of pump thrombosis in axial flow LVADs.

Prediction of aortic valve regurgitation after continuous-flow left ventricular assist device implantation using artificial intelligence trained on acoustic spectra

Significant aortic regurgitation (AR) is a common complication after continuous-flow left ventricular assist device (LVAD) implantation. Using machine-learning algorithms, this study was designed to examine valuable predictors obtained from LVAD sound and to provide models for identifying AR. During a 2-year follow-up period of 13 patients with Jarvik2000 LVAD, sound signals were serially obtained from the chest wall above the LVAD using an electronic stethoscope for 1 min at 40,000 Hz, and echocardiography was simultaneously performed to confirm the presence of AR. Among the 245 echocardiographic and acoustic data collected, we found 26 episodes of significant AR, which we categorized as “present”; the other 219 episodes were characterized as “none”. Wavelet (time–frequency) analysis was applied to the LVAD sound and 19 feature vectors of instantaneous spectral components were extracted. Important variables for predicting AR were searched using an iterative forward selection method. Seventy-five percent of 245 episodes were randomly assigned as training data and the remaining as test data. Supervised machine learning for predicting concomitant AR involved an ensemble classifier and tenfold stratified cross-validation. Of the 19 features, the most useful variables for predicting concomitant AR were the amplitude of the first harmonic, LVAD rotational speed during intermittent low speed (ILS), and the variation in the amplitude during normal rotation and ILS. The predictive accuracy and area under the curve were 91% and 0.73, respectively. Machine learning, trained on the time–frequency acoustic spectra, provides a novel modality for detecting concomitant AR during follow-up after LVAD.

Integration of novel monitoring devices with machine learning technology for scalable cardiovascular management

Ambulatory monitoring is increasingly important for cardiovascular care but is often limited by the unpredictability of cardiovascular events, the intermittent nature of ambulatory monitors and the variable clinical significance of recorded data in patients. Technological advances in computing have led to the introduction of novel physiological biosignals that can increase the frequency at which abnormalities in cardiovascular parameters can be detected, making expert-level, automated diagnosis a reality. However, use of these biosignals for diagnosis also raises numerous concerns related to accuracy and actionability within clinical guidelines, in addition to medico-legal and ethical issues. Analytical methods such as machine learning can potentially increase the accuracy and improve the actionability of device-based diagnoses. Coupled with interoperability of data to widen access to all stakeholders, seamless connectivity (an internet of things) and maintenance of anonymity, this approach could ultimately facilitate near-real-time diagnosis and therapy. These tools are increasingly recognized by regulatory agencies and professional medical societies, but several technical and ethical issues remain. In this Review, we describe the current state of cardiovascular monitoring along the continuum from biosignal acquisition to the identification of novel biosensors and the development of analytical techniques and ultimately to regulatory and ethical issues. Furthermore, we outline new paradigms for cardiovascular monitoring.

Advances in cardiovascular monitoring technologies have resulted in an influx of consumer-targeted wearable sensors that have the potential to detect numerous heart conditions. In this Review, Krittanawong and colleagues describe processes involved in biosignal acquisition and analysis of cardiovascular monitors, as well as their associated ethical, regulatory and legal challenges.

Advances in the use of cardiovascular monitoring technologies, such as the development of novel portable sensors and machine learning algorithms that can provide near-real-time diagnosis, have the potential to provide personalized care.

Wearable sensor technologies can detect numerous biosignals, such as cardiac output, blood-pressure levels and heart rhythm, and can integrate multiple modalities.

The use of novel biosignals for diagnosis raises concerns regarding accuracy and actionability within clinical guidelines, in addition to medical, legal and ethical issues.

Machine learning-based interpretation of biosensor data can facilitate rapid evaluation of the haemodynamic consequences of heart failure or arrhythmias, but is limited by the presence of noise and training data that might not be representative of the real-world clinical setting.

The use of data derived from cardiovascular monitoring devices is associated with numerous challenges, such as data security, accessibility and ownership, in addition to other ethical and regulatory concerns.