Arterial Input Function Placement for Accurate CT Perfusion Map Construction in Acute Stroke

Abstract

OBJECTIVE The objective of our study was to evaluate the effect of varying arterial input function (AIF) placement on the qualitative and quantitative CT perfusion parameters. MATERIALS AND METHODS Retrospective analysis of CT perfusion data was performed on 14 acute stroke patients with a proximal middle cerebral artery (MCA) clot. Cerebral blood flow (CBF), cerebral blood volume (CBV), and mean transit time (MTT) maps were constructed using a systematic method by varying only the AIF placement in four positions relative to the MCA clot including proximal and distal to the clot in the ipsilateral and contralateral hemispheres. Two postprocessing software programs were used to evaluate the effect of AIF placement on perfusion parameters using a delay-insensitive deconvolution method compared with a standard deconvolution method. RESULTS One hundred sixty-eight CT perfusion maps were constructed for each software package. Both software programs generated a mean CBF at the infarct core of < 12 mL/100 g/min and a mean CBV of < 2 mL/100 g for AIF placement proximal to the clot in the ipsilateral hemisphere and proximal and distal to the clot in the contralateral hemisphere. For AIF placement distal to the clot in the ipsilateral hemisphere, the mean CBF significantly increased to 17.3 mL/100 g/min with delay-insensitive software and to 19.4 mL/100 g/min with standard software (p < 0.05). The mean MTT was significantly decreased for this AIF position. Furthermore, this AIF position yielded qualitatively different parametric maps, being most pronounced with MTT and CBF. Overall, CBV was least affected by AIF location. CONCLUSION For postprocessing of accurate quantitative CT perfusion maps, laterality of the AIF location is less important than avoiding AIF placement distal to the clot as detected on CT angiography. This pitfall is less severe with deconvolution-based software programs using a delay-insensitive technique than with those using a standard deconvolution method.

Does the Location of the Arterial Input Function Affect Quantitative CTP in Patients with Vasospasm?

“…7–9 These studies revealed no significant difference in the quantitative CTP data with AIF selection ipsilateral or contralateral to the side of MCA vessel occlusion in either the ACA or MCA segments. However, AIF selection distal to an embolus revealed significant differences in the quantitative CTP data recommending avoiding its placement.…”

“…7–11 These studies found that major variations in the AIF location had no significant effect on the quantitative CTP data. 7,9–11 Additionally, the ACA and even the superficial temporal artery may be used as appropriate AIF locations in patients with acute stroke regardless of the cortical region affected which may assist in standardization of the AIF location. 7,9–11 Furthermore, there is no significant difference in the quantitative CTP data when selecting an AIF location in the MCA or ACA segments ipsilateral or contralateral to the vascular territory with infarction.…”

“…However, AIF selection distal to an embolus revealed significant differences in the quantitative CTP data recommending avoiding its placement. 9 It is unclear to what extent these findings are applicable to aSAH patients, since vasospasm results in narrowing rather than abrupt occlusion of the cerebral arteries. When significant vasospasm (≥50% narrowing) is present, perfusion deficits have been reported in the corresponding vascular territories 1,2 supporting the hypothesis that AIF selection distal to severe vasospasm could potentially affect quantitative CTP data.…”

Sources of variability in computed tomography perfusion: implications for acute stroke management

“…Arterial input function placement distal to an arterial thrombus, however, undermines perfusion calculation assumptions, and causes significant variability in CT perfusion results. 4 Two additional studies 9,19 found that varying placement of the venous output function caused significant variability in CT perfusion results. Although the anterior cerebral artery is commonly selected for the arterial input function and the superior sagittal sinus is commonly selected for the venous output function, there is no standard recommendation for selecting these user inputs.…”

“…4,13,14,22 Different corrective functions for partial-volume effects (inaccurate measurements of contrast media concentration that occur due to the small size or location of cerebral vessels) cause significantly different CT perfusion results. 25 Computed tomography perfusion software that uses a corrective technique called vascular pixel elimination was found to more closely correlate with PET results than software that does not incorporate this technique.…”