Accurate and Standardized Coronary Wave Intensity Analysis

TY - JOUR

T1 - Accurate and Standardized Coronary Wave Intensity Analysis

AU - Rivolo, Simone

AU - Patterson, Tiffany

AU - Asrress, Kaleab N.

AU - Marber, Michael

AU - Redwood, Simon

AU - Smith, Nicolas P.

AU - Lee, Jack

PY - 2017

Y1 - 2017

N2 - Objective: Coronary wave intensity analysis (cWIA) has increasingly been applied in the clinical research setting to distinguish between the proximal and distal mechanical influences on coronary blood flow. Recently, a cWIA-derived clinical index demonstrated prognostic value in predicting functional recovery post-myocardial infarction. Nevertheless, the known operator dependence of the cWIA metrics currently hampers its routine application in clinical practice. Specifically, it was recently demonstrated that the cWIA metrics are highly dependent on the chosen Savitzky-Golay filter parameters used to smooth the acquired traces. Therefore, a novel method to make cWIA standardized and automatic was proposed and evaluated in vivo. Methods: The novel approach combines an adaptive Savitzky-Golay filter with high-order central finite differencing after ensemble-averaging the acquired waveforms. Its accuracy was assessed using in vivo human data. The proposed approach was then modified to automatically perform beatwise cWIA. Finally, the feasibility (accuracy and robustness) of the method was evaluated. Results: The automatic cWIA algorithm provided satisfactory accuracy under a wide range of noise scenarios (≤10% and ≤20% error in the estimation of wave areas and peaks, respectively). These results were confirmed when beat-by-beat cWIA was performed. Conclusion: An accurate, standardized, and automated cWIA was developed. Moreover, the feasibility of beatwise cWIA was demonstrated for the first time. Significance: The proposed algorithm provides practitioners with a standardized technique that could broaden the application of cWIA in the clinical practice as well as enabling multicenter trials. Furthermore, the demonstrated potential of beatwise cWIA opens the possibility of investigating the coronary physiology in real time.

AB - Objective: Coronary wave intensity analysis (cWIA) has increasingly been applied in the clinical research setting to distinguish between the proximal and distal mechanical influences on coronary blood flow. Recently, a cWIA-derived clinical index demonstrated prognostic value in predicting functional recovery post-myocardial infarction. Nevertheless, the known operator dependence of the cWIA metrics currently hampers its routine application in clinical practice. Specifically, it was recently demonstrated that the cWIA metrics are highly dependent on the chosen Savitzky-Golay filter parameters used to smooth the acquired traces. Therefore, a novel method to make cWIA standardized and automatic was proposed and evaluated in vivo. Methods: The novel approach combines an adaptive Savitzky-Golay filter with high-order central finite differencing after ensemble-averaging the acquired waveforms. Its accuracy was assessed using in vivo human data. The proposed approach was then modified to automatically perform beatwise cWIA. Finally, the feasibility (accuracy and robustness) of the method was evaluated. Results: The automatic cWIA algorithm provided satisfactory accuracy under a wide range of noise scenarios (≤10% and ≤20% error in the estimation of wave areas and peaks, respectively). These results were confirmed when beat-by-beat cWIA was performed. Conclusion: An accurate, standardized, and automated cWIA was developed. Moreover, the feasibility of beatwise cWIA was demonstrated for the first time. Significance: The proposed algorithm provides practitioners with a standardized technique that could broaden the application of cWIA in the clinical practice as well as enabling multicenter trials. Furthermore, the demonstrated potential of beatwise cWIA opens the possibility of investigating the coronary physiology in real time.

UR - http://www.scopus.com/inward/record.url?scp=85018955715&partnerID=8YFLogxK

U2 - 10.1109/TBME.2016.2593518

DO - 10.1109/TBME.2016.2593518

M3 - Article

C2 - 27514033

SN - 0018-9294

VL - 64

SP - 1187

EP - 1196

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

IS - 5

M1 - 7530884

ER -