Format

Send to

Choose Destination
Acad Radiol. 2015 Sep;22(9):1128-37. doi: 10.1016/j.acra.2015.03.016. Epub 2015 May 30.

Coronary artery calcium: 0.5 mm slice-thickness reconstruction with adjusted attenuation threshold outperforms 3.0 mm by validating against spatially registered intravascular ultrasound with radiofrequency backscatter.

Author information

1
Department of Cardiovascular Imaging, Piedmont Heart Institute, 95 Collier Rd, Atlanta, GA 30309. Electronic address: zhen.qian@piedmont.org.
2
Department of Cardiovascular Imaging, Piedmont Heart Institute, 95 Collier Rd, Atlanta, GA 30309; Department of Computer Science, Georgia State University, Atlanta, Georgia.
3
Atlanta Medical Center, Atlanta, Georgia.
4
Department of Computer Science, Georgia State University, Atlanta, Georgia.
5
Global Genomics Group, Richmond, Virginia.
6
Department of Cardiovascular Imaging, Piedmont Heart Institute, 95 Collier Rd, Atlanta, GA 30309.

Abstract

RATIONALE AND OBJECTIVES:

Coronary artery calcium (CAC) images can be reconstructed with thinner slice thickness on some modern multidetector-row computed tomography scanners without additional radiation. We hypothesized that the isotropic 0.5-mm CAC reconstruction outperforms the conventional 3.0-mm reconstruction in detecting and quantifying coronary calcium, and we proposed to compare them by validating against spatially registered intravascular ultrasound with radiofrequency backscatter-virtual histology (IVUS-VH).

MATERIALS AND METHODS:

Twenty-seven patients were enrolled, and 5976 mm of coronary arteries were analyzed. A semiautomatic software was developed to coregister CAC and IVUS-VH on a detailed slice-by-slice basis. Calcium detection and calcium volume quantification were evaluated and compared using varying calcium attenuation thresholds. Algorithms for deriving individualized optimal threshold and comparable Agatston score on the 0.5-mm reconstruction were developed.

RESULTS:

The isotropic 0.5-mm reconstruction achieved significantly higher area under receiver-operating curve than the conventional 3.0-mm reconstruction (0.9 vs. 0.74, P < .001). Using the optimal threshold, the 0.5-mm reconstruction had higher sensitivity (0.79 vs. 0.65), specificity (0.85 vs. 0.77), positive predictive value (0.42 vs. 0.29), and negative predictive value (0.97 vs. 0.94) than the 3.0 mm. Individualized optimal threshold was significantly correlated with the image noise (r = 0.66, P < .001) in the 0.5-mm reconstruction.

CONCLUSIONS:

By optimizing the calcium threshold, the 0.5-mm reconstruction is superior to the conventional 3.0-mm in detecting and quantifying calcium, which may improve the clinical value of CAC without additional radiation.

KEYWORDS:

CT reconstruction; Coronary artery calcium; IVUS; registration

PMID:
26036721
DOI:
10.1016/j.acra.2015.03.016
[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center