Improved diagnostic performance of transluminal attenuation gradient in combination with morphological evaluation of coronary artery stenosis using 320-row computed tomography

Kyohei Nagata, Ryoichi Tanaka, Hidenobu Takagi, Tetuya Fusazaki, Yoshihiro Morino, Kunihiro Yoshioka

Research output: Contribution to journalArticlepeer-review

3 Citations (Scopus)

Abstract

Purpose: To assess the utility of transluminal attenuation gradient (TAG) in combination with coronary computed tomography angiography (CTA) for detecting obstructive coronary artery stenosis. Materials and methods: We retrospectively evaluated coronary CTA data in 35 consecutive patients who underwent invasive coronary angiography (ICA). We compared the diagnostic performance of TAG with that of quantitative coronary angiography, which we used as the reference standard. For the combination of TAG and coronary CTA, we designed a logical conjunction model (CCTA&&TAG) as well as a logical disjunction model (CCTA||TAG), and evaluated their diagnostic accuracies. Results: Among 130 vessels of 35 patients, 30 lesions (23%) showed significant stenosis on ICA. TAG predicted significant coronary artery stenosis with a sensitivity of 75%, specificity of 63%, positive predictive value of 40%, negative predictive value of 89%, and accuracy of 66%, and had a cutoff value of − 15.4 HU/10 mm. The respective values for CCTA&&TAG were 73, 98, 88, 90, and 90%; those for CCTA||TAG were 94, 61, 56, 94, and 72%. CCTA&&TAG improved the diagnostic accuracy significantly more than CCTA||TAG. Conclusion: TAG in conjunction with coronary CTA improves the diagnostic accuracy of coronary CTA.

Original languageEnglish
Pages (from-to)51-58
Number of pages8
JournalJapanese Journal of Radiology
Volume36
Issue number1
DOIs
Publication statusPublished - 2018 Jan 1
Externally publishedYes

Keywords

  • Coronary angiography
  • Coronary stenosis
  • Multidetector computed tomography

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Fingerprint

Dive into the research topics of 'Improved diagnostic performance of transluminal attenuation gradient in combination with morphological evaluation of coronary artery stenosis using 320-row computed tomography'. Together they form a unique fingerprint.

Cite this