Correcting the effects of background microcirculation in the measurement of arterial input functions using dynamic susceptibility contrast MRI of the brain

Thornton, Robert J.; Jones, Jeremy Y.; Wang, Zhiyue J.

doi:10.1016/j.mri.2005.09.014

« Previous Next »Magnetic Resonance Imaging
Volume 24, Issue 5 , Pages 619-623, June 2006

Correcting the effects of background microcirculation in the measurement of arterial input functions using dynamic susceptibility contrast MRI of the brain

Robert J. Thornton
- Department of Radiology, Baylor College of Medicine, Houston, TX 77030, USA
- Edward B. Singleton Department of Diagnostic Imaging, Texas Children's Hospital, Houston, TX 77030, USA
Jeremy Y. Jones
- Department of Radiology, Baylor College of Medicine, Houston, TX 77030, USA
- Edward B. Singleton Department of Diagnostic Imaging, Texas Children's Hospital, Houston, TX 77030, USA
Zhiyue J. Wang
- Department of Radiology, Baylor College of Medicine, Houston, TX 77030, USA
- Edward B. Singleton Department of Diagnostic Imaging, Texas Children's Hospital, Houston, TX 77030, USA
- Corresponding author. Department of Diagnostic Imaging, Texas Children's Hospital, 6621 Fannin St., MC2-2521, Houston, TX 77030, USA. Tel.: +1 832 824 5338; fax: +1 832 825 5241.

Received 6 June 2005; received in revised form 6 December 2005; accepted 6 December 2005. published online 21 February 2006.

Abstract

In dynamic susceptibility contrast MRI, the shape of the arterial input function (AIF) is commonly obtained in the near vicinity of the middle cerebral artery (MCA). However, the tissue regions where the AIF is sampled also have significant perfusion, which contributes to T₂* changes. We investigate whether correction of this effect will introduce significant changes in the measurement of the AIF and, subsequently, the assessment of the mean transit time (MTT). Clinical dynamic susceptibility data from 13 patients with brain tumors were analyzed. Patients received either single or double doses of Magnevist followed by a saline flush through a power injector. In the correction procedure, ΔR₂* was sampled in a region of gray matter approximately 1–2 cm away from the MCA and then subtracted from the ΔR₂* sampled in the immediate vicinity of the MCA. We demonstrate that in the brain, this correction of ΔR₂* due to tissue perfusion leads to a narrower width of the AIF curve obtained with ΔR₂* (mean±S.D.=7.3±2.0 and 6.4±1.7 s, before and after correction, respectively, P<.001 using a two-tailed paired t-test). Furthermore, the peak of the AIF also moved to a slightly earlier time relative to the time of arrival (mean±S.D.=4.7±0.9 and 4.3±0.8 s, before and after correction, with P<.001). With the use of the corrected AIF, the measured MTT had increased values in areas of both gray and white matter.