Regenfus Matthias, Ropers Dieter, Achenbach Stephan, Schlundt Christian, Kessler Winfried, Laub Gerhard, Moshage Werner, Daniel Werner G
Department of Internal Medicine II, Friedrich-Alexander-Universitaet Erlangen-Nuernberg, Ulmenweg 18, 91054 Erlangen, Germany.
Invest Radiol. 2003 Apr;38(4):200-6. doi: 10.1097/01.RLI.0000057030.71459.7F.
We compared maximum intensity projections (MIP) versus original source images (SI) in respect to detection of coronary artery stenoses by means of magnetic resonance (MR) coronary angiography.
MR coronary angiography was performed on 61 patients. MIP and SI were independently evaluated as to presence of significant stenoses in the proximal and midcoronary segments and compared with x-ray angiography.
A total of 315 of 427 (74%) coronary artery segments could be evaluated in MIP and 328 of 427 (77%) in SI. In segments able to be evaluated, MIP images demonstrated 84% (54/64) sensitivity and 87% (219/251) specificity, whereas SI images showed 85% (58/68) sensitivity and 90% (235/260) specificity. Overall accuracy was 87% (273/310) for MIP and 89% (293/328) for SI. There was no statistically significant difference between both modalities.
The MIP reconstructions showed comparable accuracy to unprocessed SI. However, MIP postprocessing is compromised by a higher number of images that were unable to be evaluated due to overlap of coronary arteries with adjacent cardiac structures.
