Impact of the dark rim artifact on quantification of myocardial perfusion and perfusion reserve from ssfp first pass images
- 506 Downloads
KeywordsMyocardial Perfusion Steady State Free Precession Stress Perfusion Signal Intensity Change Saturation Recovery
The measurement of myocardial perfusion by magnetic resonance imaging can be confounded by the Dark Rim Artifact (DRA). It appears as signal loss in a rim of pixels in the subendocardium at the boundary between the myocardium and LV blood. It has been shown that the DRA has similar severity in stress and rest perfusion images and occurs most frequently in images acquired with the Steady State Free Precession (SSFP) pulse sequence (1, 2).
The purpose of this study is to determine the impact of the DRA on the quantification of absolute stress and rest myocardial perfusion and perfusion reserve using the SSFP pulse sequence.
We studied 5 volunteers without know coronary artery disease (2 males) recruited with IRB consent. In each case, perfusion imaging was performed with Gadodiamide 0.05 mM/kg, at a rate of 6 ml/s and with a saturation recovery SSFP pulse sequence under adenosine (140 μg/kg) induced stress and rest conditions, with a 20-minute washout period. The temporal resolution is 160 ms and spatial resolution is 2 × 3 × 8 mm3. Three rotational long axis images (HLA, VLA, and LVOT views) were acquired. A Fermi model was used to quantify regional dynamic contrast signal derived from two sets of myocardial contours, excluding the DRA in the first instance then including the DRA in the second data set, using Medis V2006 and a custom Matlab (The MathWorks, Natick, MA) perfusion analysis program. Dynamic signal intensity changes were measured for 30%, 50% and (100%) of the myocardial as defined in Medis. The three long axis images was divided into six equal segments for analysis giving a total of 18 sectors for each volunteer.
The DRA has a significant impact on the MPR causing it to decrease. The absolute rest perfusion increases while the stress perfusion decreases only slightly relative to when the DRA is excluded form the measurement.
This article is published under license to BioMed Central Ltd.