Program Information
BEST IN PHYSICS (IMAGING): Correction of the First-Pass Distortion in Arterial Input Function of DCE MRI for Perfusion Quantification
H Wang*, T Verma , Y Cao , The University of Michigan, Ann Arbor, MI
Presentations
TH-CD-207-1 (Thursday, July 16, 2015) 10:00 AM - 12:00 PM Room: 207
Purpose:To quantify perfusion in an abdominal organ (liver, spleen or kidney) from dynamic contrast-enhanced (DCE) MRI, an arterial input function(AIF) is commonly sampled in a volume of interest (VOI) on the descending aorta. However, the peak of the first pass of the AIF could be corrupted, due to the T2* effect of high contrast-concentration, particularly at 3T MR, and other factors. This study aimed to correct this distortion in the AIF for DCE MRI perfusion quantification of an abdominal organ.
Methods:A post-processing correction method was developed to automatically recognize the distortion portion in the first pass of an AIF and fit it with a gamma-variate function. The fitting of the first pass was incorporated in pharmacokinetic(PK) modeling of the DCE curves from VOIs in both spleen and kidney. The optimal solution was determined by balancing between best fitting the PK model and matching the fitted first pass with the measured one. The corrected AIFs were applied to quantify hepatic perfusion from DCE MRI of 6 patients, which were acquired on a 3T scanner using a VIBE sequence.
Results:In the DCE-MRI of 6 patients, the original AIFs had 20%±6% saturation in the first-pass
peak compared with the corrected AIFs. Using the corrected AIFs to quantify hepatic perfusion yielded average differences of -36% (p=0.002) and 17% (p=0.01) in respective arterial perfusion and portal venous perfusion on VOIs of normal hepatic tissue, compared to those estimated using the original uncorrected AIFs.
Conclusion:The proposed method corrects the distortion in the first pass of the AIF sampled in the aorta, and could improve DCE-MRI perfusion quantification in the liver and other abdominal organs. The resultant perfusion may assess tissue function and tumor response to radiation and provide tools to support physiologically adaptive RT. The work is supported in part by RO1CA132834 and PO1CA59827
Funding Support, Disclosures, and Conflict of Interest: The work is supported in part by NIH RO1CA132834 and PO1CA59827
Contact Email: