The image quality and diagnostic performance of CT perfusion-derived CT angiography versus that of conventional CT angiography

Background: The combination of computed tomography angiography (CTA) and computed tomography perfusion (CTP) evaluation of cerebral perfusion status and vascular conditions can improve the diagnostic accuracy of infarction, ischemia, and vascular occlusion in stroke patients, as well as a comprehensive assessment of cerebral edema, collateral circulation, and blood perfusion in the lesion area. However, the consequent radiation safety and contrast agent nephropathy have aroused increasing concern. The purpose of this study was to assess the image quality and diagnostic accuracy of CTA images derived from CTP data, and to explore the feasibility of replacing conventional CTA. Methods: A total of 31 consecutive patients with suspected acute ischemic stroke were retrospectively analyzed. All patients underwent head and neck CTA and brain CTP examinations. All the CTP images were transmitted to the ShuKun artificial intelligence system, which reconstructs CTA derived from CTP (CTADF-CTP). The images were divided into 2 groups, including CTA-DF-CTP (Group A) and conventional CTA (Group B). The CT attenuation values, subjective image noise, signal-to-noise ratio (SNR), contrastto-noise ratio (CNR), image quality, CT volume dose index (CTDIvol), dose length product (DLP), and effective radiation dose (ED) were compared between the 2 groups. Moreover, the consistency of vascular stenosis and stenosis degree between the 2 groups were measured and evaluated. Results: There were no significant differences in image noise, SNR, or CNR between Groups A and B (P>0.05). The CT attenuation values of the arteries were higher in Group A than in B [ internal carotid artery (ICA) =548 +/- 112 vs. 454 +/- 85 Hounsfield units (HU), middle cerebral artery ( MCA) =453 +/- 118 vs. 388 +/- 70 HU, and basilar artery (BA) =431 +/- 99 vs. 360 +/- 83 HU] (P<0.01). The image quality of the 2 groups met the requirement of clinical diagnosis (4.97 +/- 0.18 vs. 4.94 +/- 0.25). No significant difference was found in subjective evaluation (P>0.05). In Group A compared with Group B, the following reductions were observed: CTDIvol (10.7%; 100.8 vs. 112.9 mGy), DLP ( 23.0%; 1,613 +/- 0 vs. 2,093 +/- 88 mGy center dot cm), and ED (23.0%; 5.00 +/- 0.00 vs. 6.49 +/- 0.27 mSv). Conclusions: CTA-DF-CTP data provide diagnostic accuracy and image quality similar to those of conventional CTA of head and neck CTA.