Objective:To explore the application value of reconstructed CT angiography(CTA)derived from CTP datasets by using artificial intelligence(AI)software.Methods:Sixty patients with clinically suspected ischemic cerebrovascular disease were enrolled.All cases were examined by cerebral CT,CT angiography(CTA)and CT perfusion(CTP).The CT perfusion data of all patients were imported into the brain perfusion intelligent analysis system to obtain angiography(hereafter called CTPA).We analyzed 540 segmental vessels in CTA and CTPA groups,reference as conventional CTA.The Kappa test was used to compare the consistency of image quality and cerebral artery stenosis between two groups.Results:CTPA group was similar to CTA group in terms of cerebral artery enhancement,arteriole detail display ability,image noise and venous pollution.Kw val-ues were>0.75 for the consistency test between CTPA group and CTA group on the display of cerebral artery stenosis.Only 15 segments of the 540 segmental vessels were inconsistent between CTPA and CTA.Conclusion:By using AI software,the cerebral angiography can be reconstructed conveniently and quickly from CTP datasets.The CTPA can accurately display arte-rial stenosis and occlusion compared with the CTA.