Quantification of the myocardial area at risk using coronary CT angiography and Voronoi algorithm-based myocardial segmentation

Akira Kurata, Atsushi Kono, T Sakamoto, T Kido, T Mochizuki, H Higashino, M Abe, Adriaan Coenen, Raluca Chelu, Pim Feijter, Gabriel Krestin, Koen Nieman

Research output: Contribution to journalArticleAcademicpeer-review

54 Citations (Scopus)


Objectives The purpose of this study was to estimate the myocardial area at risk (MAAR) using coronary computed tomography angiography (CTA) and Voronoi algorithm-based myocardial segmentation in comparison with single-photon emission computed tomography (SPECT). Methods Thirty-four patients with coronary artery disease underwent 128-slice coronary CTA, stress/rest thallium-201 SPECT, and coronary angiography (CAG). CTA-based MAAR was defined as the sum of all CAG stenosis (>50 %) related territories (the ratio of the left ventricular volume). Using automated quantification software (17-segment model, 5-point scale), SPECT-based MAAR was defined as the number of segments with a score above zero as compared to the total 17 segments by summed stress score (SSS), difference (SDS) score map, and comprehensive SPECT interpretation with either SSS or SDS best correlating CAG findings (SSS/SDS). Results were compared using Pearson's correlation coefficient. Results Forty-nine stenoses were observed in 102 major coronary territories. Mean value of CTA-based MAAR was 28.3 +/- 14.0 %. SSS-based, SDS-based, and SSS/SDS-based MAAR was 30.1 +/- 6.1 %, 20.1 +/- 15.8 %, and 26.8 +/- 15.7 %, respectively. CTA-based MAAR was significantly related to SPECT-based MAAR (r=0.531 for SSS; r=0.494 for SDS; r=0.814 for SSS/SDS; P<0.05 in each). Conclusions CTA-based Voronoi algorithm myocardial segmentation reliably quantifies SPECT-based MAAR.
Original languageUndefined/Unknown
Pages (from-to)49-57
Number of pages9
JournalEuropean Radiology
Issue number1
Publication statusPublished - 2015

Research programs

  • EMC COEUR-09
  • EMC NIHES-03-30-01

Cite this