Derivation and evaluation of thresholds for core and tissue at risk of infarction using CT perfusion

J Neuroimaging. 2014 Nov-Dec;24(6):562-568. doi: 10.1111/jon.12134. Epub 2014 Jul 4.

Abstract

Background and purpose: Computed tomography perfusion provides information on tissue viability according to proposed thresholds. We evaluated thresholds for ischemic core and tissue at risk and subsequently tested their accuracy in independent datasets.

Materials and methods: Tissue at risk was evaluated in patients with persistent arterial occlusions, and ischemic core thresholds in patients with recanalization and major clinical improvement. Scans were randomly allocated to derivation or validation groups for tissue at risk and core analysis. Optimum thresholds using mean transit time (MTT), cerebral blood flow (CBF), cerebral blood volume, and delay time (DT) were assessed.

Results: Absolute MTT, relative MTT and DT were best derived predictors of tissue at risk with thresholds of ≥ 7 seconds, ≥ 125%, and ≥ 2 seconds respectively. DT ≥ 2 seconds was the best predictor in the validation dataset (95% agreement levels = -44 to +30 mL, Bias = -6.9). Absolute and relative MTT were the best derived predictors of infarct volume in the core group (8 seconds and 125% respectively) but relative CBF of ≤ 45% performed best in the core validation dataset.

Conclusions: Time-based perfusion thresholds perform well as predictors of tissue at risk of infarction with DT the best predictor. Relative CBF was the best predictor of ischemic core. Evaluation in larger populations is needed to confirm the performance of tissue viability thresholds.

Keywords: Stroke; computed tomography; imaging; perfusion.

Publication types

  • Multicenter Study
  • Randomized Controlled Trial
  • Research Support, Non-U.S. Gov't
  • Validation Study

MeSH terms

  • Aged
  • Blood Flow Velocity*
  • Cerebral Angiography / methods*
  • Cerebral Infarction / diagnostic imaging*
  • Cerebral Infarction / physiopathology*
  • Cerebrovascular Circulation*
  • Female
  • Humans
  • Magnetic Resonance Angiography / methods*
  • Male
  • Reproducibility of Results
  • Risk Assessment / methods
  • Scotland
  • Sensitivity and Specificity
  • Tomography, X-Ray Computed / methods*