Format

Send to

Choose Destination
Int J Cardiovasc Imaging. 2019 Oct 18. doi: 10.1007/s10554-019-01709-3. [Epub ahead of print]

On-site evaluation of CT-based fractional flow reserve using simple boundary conditions for computational fluid dynamics.

Author information

1
Department of Cardiovascular Medicine, Kyoto University Graduate School of Medicine, 54 Shogoin-Kawahara-cho, Sakyo-ku, Kyoto, 606-8507, Japan.
2
EBM Corporation, Tokyo, Japan.
3
Biomechanics Laboratory, Electrical and Mechanical Engineering, Nagoya Institute of Technology, Nagoya, Japan.
4
Department of Cardiovascular Medicine, Gifu Heart Center, Gifu, Japan.
5
Department of Cardiovascular Medicine, Kyoto University Graduate School of Medicine, 54 Shogoin-Kawahara-cho, Sakyo-ku, Kyoto, 606-8507, Japan. naritatu@kuhp.kyoto-u.ac.jp.

Abstract

Fractional flow reserve (FFR) is an established method for diagnosing physiological coronary artery stenosis. A method for computing FFR using coronary computed tomography (CT) images was recently developed. However, its calculation requires off-site supercomputer analysis. Here, we report the preliminary result of a method using simple estimation of boundary conditions. The lumen boundaries of the coronary arteries were semi-automatically delineated using full width at half maximum of CT number profiles. The computational fluid dynamics (CFD) of the blood flow was performed using the boundary conditions of a fixed pressure at the coronary ostium and flow rates at each outlet. The total inflow at the coronary ostium was estimated based on the uniform wall shear stress hypothesis and corrected using a hyperemic multiplier to gain a hyperemic flow rate. The flow distribution from a parent vessel to the downstream daughter vessels was determined according to Murray's law. FFR estimated by CFD was calculated as FFRCFD = Pd/Pa. We collected patients who underwent coronary CT and coronary angiography followed by invasively measured FFR and compared FFRCFD with FFR. Sensitivity, specificity, and correlations were assessed. A total of 48 patients and 72 arteries were assessed. The correlation coefficient of FFRCFD with FFR was 0.56. The cut-off value was ≤ 0.80, sensitivity was 59.1%, and specificity was 94.0%. CFD-based FFR using simple boundary conditions for on-site clinical computation provided FFRCFD values that were moderately correlated with invasively measured FFR.

KEYWORDS:

Computational fluid dynamics; Computed tomography; Coronary physiology; Fractional flow reserve

PMID:
31628575
DOI:
10.1007/s10554-019-01709-3

Supplemental Content

Full text links

Icon for Springer
Loading ...
Support Center