(G-258) Hemodynamic sensitivity to heart-rate in the coronary arteries

Introduction:: Coronary artery disease (CAD) remains a significant global health concern, ranking as the third leading cause of death worldwide and the first in the United States [3]. Hemodynamic metrics, including blood flow velocity, pressure gradients, and wall shear stress, play a pivotal role in the identification of cardiovascular diseases and the prediction of disease progression [6]. These metrics serve as critical indicators of vascular functionality, helping in the diagnosis and treatment of cardiovascular diseases. While traditional analysis has focused on metric values under specific conditions, such as resting or hyperemic states, the rise of continuous monitoring devices, such as wearables, has sparked interest in understanding the sensitivity of these metrics to changes in heart rate. Current methods mainly rely on isolated snapshots of a patient’s hemodynamic state, generating essential metrics such as fractional flow reserve for treatment guidance [12][11][13]. However, a comprehensive view of the hemodynamics of a patient over time could significantly improve diagnostics and disease tracking. Computational fluid dynamics (CFD) offers a non-invasive means to access a patient’s hemodynamic states [8], but its computational expense necessitates understanding the sensitivity of these metrics to flow condition changes and determining the required number of simulations. This study presents a systematic investigation of the sensitivity of various hemodynamic metrics to changes in heart rate. By isolating these sensitivities, our goal is to advance our understanding of CAD progression and provide valuable information to improve diagnostic accuracy and treatment efficacy.

Materials and Methods::

Results, Conclusions, and Discussions:: Figure 1 shows heatmaps displaying pairwise comparisons of heart rate. In the near-outlet location, the pressure gradient, velocity, and WSS have errors of 29%, 71%, and 76%, respectively, between 50 and 200 bpm with 200 bpm as ground truth. Between 50-60 bpm, the same parameters have errors of 7.5%, 15%, and 17%, while between 190-200 bpm, the errors are 4.8%, 5.1%, and 6.9%. With a threshold of 1% percent error, all indicate a need for more granular step sizes. WSS is the most sensitive to heart rate, which is supported by error propagation. LBM pressure, calculated from the LBM density, is relatively insensitive to changes in heart rate, as the density is derived from hematocrit, not flow rate [7]. The asymmetry found in some heat maps emphasizes the higher sensitivity corresponding to lower heart rates, suggesting flexibility in heart rate variability at higher activity levels. This is supported by the slight fanning out towards higher heart rate and percent error doubling at lower heart rate. The ostial location has higher percent errors for WSS and velocity than at the outlet, as near-outlet regions may experience a damping effect from branching or from distance to the ostium. The pressure gradient heat maps in Fig. 1-B show a bifurcation at 130-140 bpm, the parameter exhibits a sharp increase at 140 bpm, requiring further investigation for an exact threshold. Near the outlet, the pressure gradient is almost five times more sensitive than the prestenosis location, as the stenosis is closer to the ostium in this patient. As such, the bar graphs in Fig. 1C indicate a high variance for the pressure gradient. While further granularity is certainly needed, these results convey sensitivity between (pairwise comparisons), across (wide range of rates) and against (sweep of ground truths) different heart rate. It is critical to consider dissimilarity in the hemodynamic response when using patient-varying data, particularly for clinical applications. The results reveal that, especially at lower heart rates and for pressure gradient metrics, continuous biomarker data are more representative of a patient’s hemodynamic activity. The data support hemodynamic tracking beyond resting and hyperemic cases.

Acknowledgements (Optional): :

References (Optional): :