Reproducibility of ABC/2 method to determine infarct volume and mismatch percentage with CT perfusion

J Neuroimaging. 2014 May-Jun;24(3):232-7. doi: 10.1111/jon.12001. Epub 2013 Jan 16.

Abstract

Background: Our aim is to implement a simple, rapid, and reliable method using computed tomography perfusion imaging and clinical judgment to target patients for reperfusion therapy in the hyper-acute stroke setting. We introduce a novel formula (1-infarct volume [CBV]/penumbra volume [MTT] × 100%) to quantify mismatch percentage.

Methods: Twenty patients with anterior circulation strokes who underwent CT perfusion and received intravenous tissue plasminogen activator (IV tPA) were analyzed retrospectively. Nine blinded viewers determined volume of infarct and ischemic penumbra using the ABC/2 method and also the mismatch percentage.

Results: Interrater reliability using the volumetric formula (ABC/2) was very good (intraclass correlation [ICC] = .9440 and ICC = .8510) for hemodynamic parameters infarct (CBV) and penumbra (MTT). ICC coefficient using the mismatch formula (1-MTT/CBV × 100%) was good (ICC of .635).

Conclusions: The ABC/2 method of volume estimation on CT perfusion is a reliable and efficient approach to determine infarct and penumbra volumes. The 1-CBV/MTT × 100% formula produces a mismatch percentage assisting providers in communicating the proportion of salvageable brain and guides therapy in the setting of patients with unclear time of onset with potentially salvageable tissue who can undergo mechanical retrieval or intraarterial thrombolytics.

Keywords: Stroke; computed tomography; thrombolytics.

MeSH terms

  • Algorithms*
  • Cerebral Angiography / methods*
  • Cerebral Infarction / diagnostic imaging*
  • Cerebral Infarction / etiology
  • Female
  • Humans
  • Imaging, Three-Dimensional / methods*
  • Infarction, Anterior Cerebral Artery / complications
  • Infarction, Anterior Cerebral Artery / diagnostic imaging*
  • Male
  • Middle Aged
  • Observer Variation
  • Radiographic Image Enhancement / methods
  • Radiographic Image Interpretation, Computer-Assisted / methods*
  • Reproducibility of Results
  • Sensitivity and Specificity
  • Single-Blind Method
  • Tomography, X-Ray Computed / methods*