Morphological Observation And Comparative Analysis Of Willis Ring In Patients With Acute Cerebral Infarction By MRA And DSA

Objective To analyze the morphological characteristics of Willisrings in patients with acute cerebral infarction by using magneticresonance angiography (MRA)； Considering the findings of digitalsubtraction angiography (DSA) as the standard, this study intends toexplore the imaging diagnostic value of MRA in showing the cerebralfeeding arterial lesions in patients with acute cerebral infarction.Methods A retrospective analysis was carried out on the imaging dataof203patients diagnosed as acute cerebral infarction by simultaneousMRA and DSA examinations. According to Krabbe-Hartkamp classificationcriteria,the MRA morphological characteristics of Willis were observedin patients with acute cerebral infarction.Considering the findings ofDSA as the gold standard, to calculate the sensitivity, specificity,accuracy,a false positive rate,a false negative rate and Youden indexof of MRA in diagnosis of cerebral feeding arterial lesions (stenosis≥50%or occlusion) and compare the consistency between DSA and MRA,and evaluate the imaging value of MRA in diagnosis of cerebral feedingarterial lesions (stenosis≥50%or occlusion).Results In this study,among203Willis rings in patients with acutecerebral infarction,there were108cases of complete anterior Willisrings(a~f,53.21%)and95cases of incomplete anterior Willis rings(g~j, 46.79%)；there were45cases of complete posterior Willis rings(a~c,22.17%)and158cases of incomplete posterior Willis rings(d~j,77.83%).Among ten subtypes,the most common anterior ring variation was a-type(normal type),followed by h-type (hypoplasia or absence of unilateralACA-A1segment)；the most common posterior ring variation was e-type(hypoplasia or absence of bilateral PCoA),followed by a-type(normaltype)and d-type(hypoplasia or absence of unilateral PCoA).Among203patients underwent MRA diagnosis,there were299cases of vascularlesions (stenosis≥50%or occlusion),including94cases of MCA(31.44%),72cases of ICA intracranial segment(24.08%) and55cases ofVA intracranial segment(18.39%).Among203patients underwent DSAdiagnosis, there were256cases of vascular lesions (stenosis≥50%orocclusion), including87cases of MCA(33.98%),65cases of ICAintracranial segment(25.38%) and48cases of VA intracranialsegment(18.75%). Considering the results of DSA as the standards, MRAdiagnosis of cerebral artery lesions (stenosis≥50%or occlusion) waswith a sensitivity of87.89%%,a specificity of96.26%,an accuracy of95.30%,a false positive rate of3.74%,a false negative rate of12.11%，Youden index of84.15%，k=0.784；MRA diagnosis of internal carotid arteryintracranial segment lesions (stenosis≥50%or occlusion) was with asensitivity of93.85%,a specificity of96.77%,an accuracy of96.31%,afalse positive rate of3.23%,a false negative rate of6.15%,Youden indexof90.62%，k=0.868；MRA diagnosis of anterior cerebral artery lesions(stenosis≥50%or occlusion) was with a sensitivity of84.00%,aspecificity of97.90%,an accuracy of97.04%,a false positive rate of2.10%,a false negative rate of16.00%,Youden index of81.9%，k=0.762； MRA diagnosis of middle cerebral artery lesions (stenosis≥50%orocclusion) was with a sensitivity of98.85%,a specificity of97.49%,anaccuracy of97.78%,a false positive rate of2.51%,a false negative rateof1.15%,Youden index of96.34%，k=0.936;MRA diagnosis of posteriorcerebral artery lesions (stenosis≥50%or occlusion) was with asensitivity of61.5%,a specificity of96.44%,an accuracy of95.32%,afalse positive rate of4.54%,a false negative rate of38.5%,Youden indexof57.49，k=0.435;MRA diagnosis of basilar artery lesions (stenosis≥50%or occlusion) was with a sensitivity of66.67%,a specificity of91.89%,an accuracy of89.66%,a false positive rate of8.11%,a falsenegative rate of33.33%,Youden index of58.56%，k=0.478;MRA diagnosisof vertebral artery lesions (stenosis≥50%or occlusion) was with asensitivity of77.08%,a specificity of88.50%,an accuracy of94.97%,afalse positive rate of11.50%,a false negative rate of22.92%,Youdenindex of65.58%，k=0.678.Conclusion The incidence rate of Willis ring variation is higherin patients with acute cerebral infarction and common in the posteriorring variation.As a non-invasive, economical, convenient and reliablecerebrovascular imaging technique, MRA plays an increasingly role inscreening and diagnosis of cerebrovascular diseases. MRA has a highsensitivity, specificity and accuracy in the diagnosis of intracranialICA and MCA lesions (stenosis≥50%or occlusion), which is wellconsistent with DSA. Therefore, MRA can be an alternative of DSA to someextent as a routine examination to assess intracranial ICA and MCAcerebrovascular lesions (stenosis≥50%or occlusion). However, MRAdiagnosis of PCA, BA and VA posterior circulation intracranial lesions (stenosis≥50%or occlusion) has a poor sensitivity, specificity andconsistency. Therefore, to determine the nature of the lesions, a directDSA examination is recommended for patients with clinically highlysuspected posterior circulation intracranial stenosis.