Key result
Novel pump speed waveform analysis detects aortic valve opening with ~95% sensitivity vs echocardiography.
Why the study?
Instantaneous flow estimation to detect aortic valve opening in patients with LVAD support is difficult, especially in axial pumps, necessitating a new detection method.
Can pump speed waveform analysis accurately detect aortic valve opening and its duration in patients on LVAD support compared to echocardiography?
Observational (n=28)
Can pump speed waveform analysis accurately detect aortic valve opening and its duration in patients on LVAD support compared to echocardiography?
Pump speed waveform analysis using AUC of power spectral density can accurately detect aortic valve opening status and duration in LVAD patients, though patient-specific calibration is required.
May enable noninvasive aortic valve opening monitoring in LVAD patients; leaves open prospective outcome validation before clinical adoption.
As the aortic valve (AV) opens, the pump pressure head remains constant, which is reflected as a "notch/plateau" in pump pressure and flow signals. However, instantaneous flow estimation may be influenced by friction and is particularly difficult in axial pumps. Therefore, a new method to determine the duration of AV opening based on the area under the curve (AUC) of the power spectral density analysis of pump speed signal was developed. Data from patients implanted with HeartWare HVAD left ventricular assist device were studied at different pump speeds, with simultaneous transthoracic echocardiography in two cohorts. In the first group, pump data of 15 patients were used to investigate the ability to discriminate between an open and closed AV. In the second cohort of a further 13 patients, the duration of AV opening was measured from digitized M-mode images, and the relationship between the AV opening time and the new method assessed. In 14 of the initial 15 patients, AV status could be discriminated using only one threshold for all patients. In the second cohort, gradual speed reduction resulted in aortic valve opening in 12 of the 13 patients. The correlation between AV opening duration and AUC was 0.96 ± 0.03. Regression analysis indicated a linear relationship in each of the patients with a small error between the fit and the measured opening time (root mean square error = 11.0 ± 7.6 ms). However, the slopes (69.0 ± 52.8) and intercepts (-31.4 ± 78.0) varied widely between patients. The sensitivity and specificity for the new method using AUC threshold of 0.95 for aortic valve closure was 95% and 91%, respectively. The newly developed method to detect AV opening not only provides information on the AV status during LVAD support (open/closed) but also gives insight into the duration of AV opening. Because the slope of the relationship varies from patient to patient, initial training and adaptation of the method to each patient seems to be required.
No takes yet. Share an insight, caveat, or question.
Hayward et al. (2015) conducted an observational in HeartWare HVAD left ventricular assist device support (n=28). Area under the curve (AUC) of the power spectral density analysis of pump speed signal vs. Transthoracic echocardiography was evaluated on Duration of aortic valve opening and discrimination of open/closed status. A novel method using the area under the curve of pump speed power spectral density detected aortic valve opening with 95% sensitivity and 91% specificity, correlating highly with echocardiography.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: