Combined diffusion weighting and CSF suppression in functional MRI.
Författare
-
Linda Knutsson
-
Max Bolling
-
Ronnie Wirestam
-
Stig Holtås
-
Freddy Ståhlberg
Summary, in English
In this study, EPI pulse sequences with diffusion weighting for reduction of contributions from large vessels and inversion pulses in order to minimize the effects of CSF pulsations and CSF partial volume effects were developed for BOLD contrast investigations in functional MR imaging. One inversion recovery echo-planar imaging (IR-EPI) pulse sequence and one IR-EPI with additional diffusion weighting (DW-IR-EPI) were developed and compared to a standard gradient-echo EPI sequence in a cortical stimulation experiment in nine healthy volunteers. Stimulation of motor cortex was performed using a semi-complex finger-tapping paradigm in seven periods of alternating rest and stimulation. Comparison between the three pulse sequences was made by measuring the activated volume in each subject, as well as by calculating the relative signal increase during stimulation. Due to different baseline signal-to-noise levels in the images generated by the three pulse sequences, artificial noise was added so that the comparative investigation could be performed independently of the noise level. The activated volume was 128 +/- 73 pixels (mean +/- SD) using the standard EPI pulse sequence, 31 +/- 12 pixels using IR-EPI and 15 +/- 13 pixels when DW-IR-EPI was employed. The relative signal increase was 5.7 +/- 1.1% using standard EPI, 11.5 +/- 3.1% using IR-EPI and 9.9 +/- 2.4% using DW-IR-EPI. The activated volume obtained with the addition of extra noise, i.e. at equal S/N, was 70 +/- 50 pixels using the standard EPI, and when using IR-EPI, the activated volume was 28 +/- 13 pixels. At equal S/N, the signal increase was 7.3 +/- 1.4% using standard EPI and 12.0 +/- 3.6% using IR-EPI. In BOLD contrast imaging, a combination of diffusion weighting and inversion recovery appeared to reduce false activation caused by CSF pulsation and blood flow in large vessels.