Objective:

1. Comparison of conventional CT perfusion (CTP) threshold (CBF <30% and Tmax > 6 seconds) and modified CTP threshold (CBF <38%, and Tmax > 4 seconds) with diffusion weighted imaging (DWI) positive strokes on MRI for detection of acute ischemic stroke.
2. Evaluate their sensitivity and specificity in detecting stroke using MRI DWI positivity as the gold standard. 

Background:

Common parameters used in CTP scans have been CBF and Tmax. The thresholds traditionally used for the CTP parameters have been CBF <30% and Tmax >6 seconds. Setting the threshold at CBF <38% and Tmax >4 seconds also detected acute ischemic strokes, but this has not been validated in the literature.

Design/Methods:

IRB waiver was obtained. 76 patients who underwent CT perfusion scan and MRI brain for acute stroke evaluation during a seven month period in a teaching hospital were screened. SPSS statistical software was used. Sensitivity and specificity was calculated. Cohen’s kappa analysis was done among the two thresholds.

Results:

Considering DWI on MRI for diagnosis of acute stroke as the gold standard, conventional CTP threshold had a lower sensitivity (54%) and higher specificity (85%) while modified CTP threshold had a higher sensitivity (96%), but lower specificity (23%).

Cohen's κ was calculated to determine if there was agreement among the diagnoses of stroke using modified and conventional CTP thresholds. There was poor agreement between the two judgments, κ = 0.142 (p=0.045).

Conclusions:

The modified CTP threshold proposed has a higher sensitivity but a lower specificity to detect acute ischemic stroke when compared to the conventional criteria.