Abstract: Objective To assess differences in cerebral blood flow perfusion and risk factors for clinical symptoms in two groups of patients with M1 segment of unilateral middle cerebral artery(MCA) stenosis with and without clinical symptoms by using three-dimensional pseudo-continuous arterial spin labeling(3D pCASL) imaging with two post-labeling delay(PLD) and Rapid software analysis of magnetic resonance perfusion-weighted imaging(PWI). 
Methods A total of 101 patients with M1 segment of unilateral MCA stenosis who underwent 3D pCASL with two PLD and Rapid soft ware analysis of magnetic resonance perfusion imaging(PWI). Binary logistic regression model was established, the presence or absence of symptoms was used as a binary dependent variable, and the local arterial stenosis rate, early-arriving flow proportion(EAFP),late-arriving retrograde flow proportion(LARFP), hypoperfusion intensity ratio(HIR) of downstream of the stenotic artery, and the patient′s age, gender, smoking history, homocysteine, blood pressure, blood lipids and blood glucose parameters were included in the model for analysis as independent variables. 
Results Univariate comparison of exposure factors in the symptomatic and asymptomatic group showed significant differences in LARFP and HIR between two groups, and logistic regression analysis of risk factors in patients with symptomatic stenosis after adjusting for EAFP, arterial stenosis rate, and other parameters showed that LARFP and HIR, which represented collateral blood flow, were independent risk factors for symptomatic stenosis. 
Conclusion Collateral blood flow indownstream of intracranial arterial stenosis is the determinant of clinical symptoms.

Key words: middle cerebral artery, perfusion index, collateral circulation