Article Text

Download PDFPDF
P-028 Quantitative MR perfusion and validation against stable-isotope microspheres
  1. NS Saadat1,
  2. Y Jeong2,
  3. T Carroll1,
  4. K Kawaji3,
  5. S Roth4,
  6. G Christoforidis1
  1. 1Radiology, University of Chicago, Chicago, IL
  2. 2Biomedical Engineering, Northwestern University, Chicago, IL
  3. 3Biomedical Engineering, Illinois Institute of Technology, Chicago, IL
  4. 4University of Illinois at Chicago, Chicago, IL


Purpose This study sought to assess the accuracy of a quantitative MRI-based (qMRI) measure of cerebral blood flow (qCBF) against reference stable isotope neutron capture microsphere based cerebral blood flow quantification in an experimental model during normocapnia, hypercapnia and middle cerebral artery (MCAO).

Materials/methods Five female, mongrel dogs (20–30 kg) were each studied over two days. On day 1, qCBF images were acquired during normocapnia (target PaCO2 range 30–40 mmHg) and during hypercapnia (target PaCO2 range >60 mmHg) induced by carbogen gas inhalation (5% CO2/95% O2). On day 2 animals underwent angiographically verified permanent endovascular occlusion of the M1 segment while in a normocapnic state. Anesthesia was selected to minimize influence on cerebrovascular reactivity. Physiologic parameters were maintained within normal range with the exception of PCO2 for hypercapnia. Neutron capture microspheres were injected at the time of qMRI acquisition to obtain reference-standard CBF values. qMRI was acquired on a 3 Tesla unit (Achieva, Philips Healthcare, Best, Netherlands) using a 15-channel receive-only head coil a ‘bookend dynamic susceptibility (DSC)’ approach, which uses pre- and post-contrast T1 maps bookended to a DSC MRI sequence to calculate parenchymal T1 changes and calibrate the DSC scan for quantitative perfusion in ml/100 g/min. T1 maps are derived using 2D EPI Look-Locker inversion recovery (FOV/Matrix=220 mm/224, Slice Thickness=4 mm, single slice) with variable delay time and DSC perfusion (FOV/Matrix=20 mm/224, single shot, EPI, fat saturated, slice thickness=6 mm, TR/TE=315/40, 200 time points). A gadolinium-based contrast agent (Multihance, Bracco, Princeton, NJ, USA) was mechanically injected followed by a saline flush.

Results MRI correlated strongly with microsphere perfusion (qCBFMRI =0.93*qCBFSPHERES +3.85 ml/100 g/min; r2 =0.96; p<0.001), for individual CVR (CVRMRI =1.17*CVRSPHERES – 0.95%; CBF/mmHg CO2; r2 =0.84; p<0.001), and for post-occlusion CBF (qCBFMRI =0.80*qCBFSPHERES +12.9 ml/100 g/min; r2 =0.82; p=0.002). Correction for delay and dispersion resulted in a significant improvement in the correlation between MRI and microsphere deposition in the ischemic state (qCBFMRI =0.97*qCBFSPHERES +2.58 ml/100 g/min; r2 =0.96; p<0.001).

Conclusion MRI derived values of CBF are strongly correlated with reference value microsphere deposition in normocapnia, hypercapnia, and MCAO ischemic stroke. Correction for delay and dispersion significantly improved the accuracy of this quantification during MCAO, underscoring the importance for this correction under focal ischemic condition.

Abstract P-028 Figure 1

Correlation plots of MR-qCBF vs microsphere-qCBF with MCAO (A) before and (B) after delay and dispersion correction. Black dots and dashed lines represent hemispheric averages; regional averages are gray. An improvement in the correlation can be seen after the correction. A line of is shown for reference

Disclosures N. S. Saadat: None. Y. Jeong: None. T. Carroll: None. K. Kawaji: None. S. Roth: None. G. Christoforidis: None.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.