5780 Comparison of breathhold versus PACE triggering in renal BOLD MRI

Comparison of breathhold versus PACE triggering in renal BOLD MRI Maryam Seif, Chris Boesch, and Peter Vermathen Depts Clinical Research and Radiology, University of Bern, Bern, Bern, Switzerland Background: The estimation of renal tissue oxygenation can play an important role in investigating renal dysfunctions1. Blood oxygenation level dependent (BOLD) MRI is a non-invasive approach that can monitor renal pO2 by measuring R2* (= 1 / T2*) 2. Previous studies have shown the potential of BOLD to investigate healthy and transplanted kidney1-8. To minimize motion artifacts caused by respiratory motion, BOLD kidney MRI has been performed in breath hold, which may not be feasible for patients6. At least in those patients that cannot hold their breath long enough, respiratory triggered renal BOLD measurement would be beneficial and facilitate the abdominal measurements. Purpose: The current study therefore is aimed at evaluating the feasibility of the Prospective Acquisition CorrEction (PACE) triggered renal BOLD MRI in comparison with breath-hold (BH) BOLD MRI. Methods: The study was approved by the local ethics committee and written informed consent was obtained from all subjects. Eleven healthy volunteers (6 female, 5 male, age =29.1±9.5y) without a history of chronic diseases underwent a multiple-gradient recalled echo (mGRE) BOLD sequence within a single breath-hold as well as with the PACE triggering technique on a clinical 3T MRI scanner (Trio, Siemens Erlangen Germany). The BH-BOLD sequence was separately performed 3 times to acquire 3 coronal slice positions with TR of 65 msec. Twelve T2* weighted images corresponding to 12 echoes were obtained within one 17 second breath-hold for each slice position. The PACE-BOLD sequence was performed with 5 coronal slice positions. Average duration for the PACE-BOLD was 5.7±1.5 min. The remaining parameters for breath-hold and PACE triggered BOLD sequences were as follows: 12 TEs of 6-52.30 msec, Flip Angle=300, BW= 330 Hz/Px, FOV=400 x 400 mm2, Matrix 256×256, slice thickness of 5 mm. Processing: In order to further reduce potential respiratory motion artifacts, registration of individual images was performed using an in-house developed image co-registration software based on the method proposed by H. Lu et al.9. At least six regions of interest (ROIs) were placed on 3 slices for each subject in medulla and cortex on BHBOLD images with individual ROI sizes of 90±17 and 77±13 pixels, respectively as well as on PACE-BOLD scans with individual ROI sizes of 100±16 and 76±14 pixels, respectively. Two subjects were excluded from the analysis, due to the poor visual quality of PACE triggered scans and because significantly fewer ROIs could be analyzed than on the corresponding BH-BOLD images. The R2* maps were generated by fitting the signal intensity vs. TE data to a linear function. Results: The T2* weighted images and R2* maps acquired in BHBOLD visually demonstrated less distortion, compared to the PACEBOLD MRI for most subjects (Fig.1). Table 1 shows that the mean R2* values of medulla and cortex were similar for breath-hold and PACE-BOLD scans. Mean R2* in medulla and cortex was highly significantly different in both breath-hold and PACE-BOLD measurements (P<0.001). However, cortical R2* was slightly, but significantly higher in PACE-BOLD than in BH-BOLD, while they were very similar in medulla. There was a significant correlation between Breath-hold and PACE-BOLD of medullary and cortical R2* values (Medullary R2 = 0.69, P=0.005 and Cortical R2 = 0.61, P=0.01, Fig.2). Discussions & Conclusions: Our evaluations of R2* demonstrate that PACE triggering for BOLD measurements is clearly feasible in renal investigations and yields comparable results to the BH-BOLD approach. In contrast to the visual differences between BH& PACE-BOLD images, the quantitative analysis of R2* yielded similar results for both methods, despite the slightly higher cortical R2* in PACE-BOLD. R2* values of both methods are somewhat smaller than previously reported, however, substantially different values have also been reported before8, possibly demonstrating the sensitivity of BOLD MRI to scan parameters and shimming status. In conclusion, the small difference of means and the correlations between BH-BOLD and PACE-BOLD R2* values suggest that PACEBOLD MRI can facilitate renal measurements on elderly and diseased subjects, who may have problems holding their breath. Fig. 1: (A) BH-BOLD images, (B) PACE BOLD images from the shortest & longest two echo times and corresponding R2* maps. Medullary mean R2* (s-1 ) Cortical mean R2*(S-1) BH BOLD MRI 23.6 ±2.0 17.9±1.1 PACE BOLD MRI 23.4±1.1 18.9±1.4 t-test p=0.5 p=0.01 Table 1: Comparison of mean R2* between BH and PACE BOLD