Elsevier

Academic Radiology

Volume 17, Issue 1, January 2010, Pages 61-66
Academic Radiology

Original investigation
Increasing Sampling Interval in Cerebral Perfusion CT: Limitation for the Maximum Slope Model

https://doi.org/10.1016/j.acra.2009.07.009Get rights and content

Rationale and Objectives

The aim of this study was to evaluate increased sampling intervals on cerebral dynamic perfusion computed tomographic (PCT) imaging calculated using software relying on the maximum slope model.

Materials and Methods

PCT data sets from 32 patients with suspected acute stroke were acquired with a sampling interval of 1 image/s. The PCT data sets were modified to simulate sampling intervals of 2, 3, and 4 seconds. Maps of cerebral blood flow (CBF), cerebral blood volume, and time to peak (TTP) were calculated using software relying on the maximum slope model. Parenchymal and vascular peak enhancement; absolute values of CBF, cerebral blood volume, and TTP in the nonischemic hemisphere; and ischemic area in the different perfusion maps were measured.

Results

Parenchymal peak enhancement of the nonischemic hemisphere was statistically significantly decreased in all simulated data sets with >1-second sampling intervals (P < .001). Absolute CBF and TTP values in the nonischemic hemisphere were increased in all simulated data sets with >1-second sampling intervals (P = .044–.001 and P = .008–.001, respectively). The ischemic area was significantly underestimated for CBF and TTP in all simulated data sets with >1-second sampling intervals (P = .022–.005 and P = .019–.005, respectively).

Conclusions

Sampling intervals of >1 second on PCT imaging calculated using software relying on the maximum slope model significantly alter absolute CBF and TTP values and the size of ischemia in CBF and TTP. Thus, increasing the sampling interval on dynamic PCT imaging cannot be recommended in combination with this algorithm.

Section snippets

Overview

From our stroke database, PCT examinations of 32 adult patients (18 women; mean age, 63.3 ± 16.1 years) with suspected acute stroke of the anterior circulation and symptom onset of <6 hours were chosen retrospectively. All patients underwent multimodal CT imaging according to our institutional guidelines for acute stroke, including nonenhanced computed tomography, perfusion computed tomography, and CT angiography. All patients had follow-up with nonenhanced CT imaging or diffusion-weighted

Qualitative Analysis

The visual assessment of the color-coded perfusion maps revealed diagnostic quality in all perfusion maps calculated from PCT data sets (32 of 32) with 1-second and 2-second sampling frequencies. The PCT data sets with a simulated sampling frequency of 3 seconds resulted in perfusion maps of nondiagnostic quality for 2 of 32 data sets, whereas a sampling interval of 4 seconds revealed nondiagnostic quality of the calculated perfusion maps for 9 of 32 data sets. An example is given in Figure 1.

Quantitative Analysis

Discussion

Commercial software for the processing of PCT data, on the basis of different algorithms, is available (16). Discordant findings have been reported for the postprocessing of dynamic PCT data with increased sampling intervals. Wintermark et al (13) and Wiesmann et al (14) demonstrated that PCT data sets with increased sampling intervals up to 4 seconds are feasible for postprocessing with software relying on the central-volume principle, while Kämena et al (15) found that PCT data sets with

Conclusions

PCT imaging can be safely performed with acceptable radiation exposure when fundamental aspects of the scan protocol are considered. However, PCT calculation using postprocessing software relying on the maximum slope model is significantly inferior for dynamic PCT data sets with sampling intervals >1 second compared to the original data sets with sampling intervals of 1 second. Although the visual assessment of the color-coded perfusion maps revealed no significant differences between simulated

References (21)

There are more references available in the full text version of this article.

Cited by (0)

View full text