Arterial input function placement effect on computed tomography lung perfusion maps

Quantitative Imaging in Medicine and Surgery

Chris Dey¹  Hatem Mehrez³  Anastasia Oikonomou¹  Fatima Ursani²  Narinder Paul⁴  Laura Jimenez-Juan¹  Shabnam Homampour⁴ 

[1] Department of Medical Imaging, University of Toronto, Toronto, Ontario, Canada; Department of Medical Imaging, Sunnybrook Health Science Centre, Toronto, Ontario, Canada;;Department of Biology, University of Toronto, Toronto, Ontario, CanadaToshiba Medical Systems, Markham, Ontario, Canada;Department of Medical Imaging, University of Toronto, Toronto, Ontario, Canada;Department of Medical Imaging, University Health Network, Toronto, Ontario, Canada;

关键词: Perfusion imaging;    computed tomography (CT);    pulmonary artery;   

DOI  :  10.3978/j.issn.2223-4292.2016.01.05

学科分类：外科医学

来源: AME Publications

【 摘 要 】

Background: A critical source of variability in dynamic perfusion computed tomography (DPCT) is the arterial input function (AIF). However, the impact of the AIF location in lung DPCT has not been investigated yet. The purpose of this study is to determine whether the location of the AIF within the central pulmonary arteries influences the accuracy of lung DPCT maps.

Methods: A total of 54 lung DPCT scans were performed in three pigs using different rates and volumes of iodinated contrast media. Pulmonary blood flow (PBF) perfusion maps were generated using first-pass kinetics in three different AIF locations: the main pulmonary trunk (PT), the right main (RM) and the left main (LM) pulmonary arteries. A total of 162 time density curves (TDCs) and corresponding PBF perfusion maps were generated. Linear regression and Spearman’s rank correlation coefficient were used to compare the TDCs. PBF perfusion maps were compared quantitatively by taking twenty six regions of interest throughout the lung parenchyma. Analysis of variance (ANOVA) was used to compare the mean PBF values among the three AIF locations. Two chest radiologists performed qualitative assessment of the perfusion maps using a 3-point scale to determine regions of perfusion mismatch.

Results: The linear regression of the TDCs from the RM and LM compared to the PT had a median (range) of 1.01 (0.98�?1.03). The Spearman rank correlation between the TDCs was 0.88 (P<0.05). ANOVA analysis of the perfusion maps demonstrated no statistical difference (P>0.05). Qualitative comparison of the perfusion maps resulted in scores of 1 and 2, demonstrating either identical or comparable maps with no significant difference in perfusion defects between the different AIF locations.

Conclusions: Accurate PBF perfusion maps can be generated with the AIF located either at the PT, RM or LM pulmonary arteries.

【 授权许可】

Unknown