Effects of window and threshold levels on the accuracy of three-dimensional rendering techniques in coronary artery electron-beam CT angiography

Citation
B. Lu et al., Effects of window and threshold levels on the accuracy of three-dimensional rendering techniques in coronary artery electron-beam CT angiography, ACAD RADIOL, 8(8), 2001, pp. 754-761
Citations number
29
Categorie Soggetti
Radiology ,Nuclear Medicine & Imaging
Journal title
ACADEMIC RADIOLOGY
ISSN journal
10766332 → ACNP
Volume
8
Issue
8
Year of publication
2001
Pages
754 - 761
Database
ISI
SICI code
1076-6332(200108)8:8<754:EOWATL>2.0.ZU;2-C
Abstract
Rationale and Objectives. The authors performed this study to evaluate the effect of window level and gray-scale threshold on the demonstration of cor onary artery lumina at three-dimensional electron-beam computed tomographic (CT) angiography. Materials and Methods. Forty-four coronary artery branches in postmortem pi gs were evaluated with electron-beam CT angiography, and the findings were compared with those from conventional angiography. Images from electron-bea m CT angiography were reconstructed with maximal intensity projection (MIP) , multiplanar reformation (NMR), and shaded-surface display (SSD). Four cat egories of window level and gray-scale threshold were evaluated. Results. Three-dimensional electron-beam CT angiography accurately depicted the luminal diameters of the coronary arteries compared with conventional angiography (r = 0.83-0.90, P < .0001). The length of lumina visualized at electron-beam CT angiography was significantly shorter than that visualized with conventional angiography (P < .001). The use of NUR enabled visualiza tion of longer segments of coronary arteries than did the use of MIP or SSD (P < .05). The higher the window level and gray-scale threshold used, the smaller the coronary luminal diameters measured (P < .05). The most accurat e window level and gray-scale threshold (82.6 HU +/- 29.8 and 89.5 HU +/- 2 9.7, respectively) were found to correspond to the attenuation of the lumin a (275.8 HU +/- 58.8). Results of simple linear regression showed a strong correlation between luminal attenuation and window level (r = 0.89, P < .00 01) or gray-scale threshold (r = 0.95, P < .0001). Conclusion. Electron-beam CT angiography shows promise in the visualization of coronary artery lumina. For accurate display of lumina, a proper window level and gray-scale threshold for three-dimensional rendering techniques should be determined and used on the basis of the attenuation of the target vessel.