OBJECTIVE: The overall accuracy of transcranial Doppler sonography in the diagnosis of middle cerebral artery spasm has not been established. Moreover, the factors of age and sex have not been addressed in most studies. In this article, we present receiver operating characteristic (ROC) curve analysis of the accuracy of transcranial color Doppler sonography in diagnosing middle cerebral artery spasm on the basis of flow velocities standardized for age and sex. SUBJECTS AND METHODS. We prospectively studied 214 consecutive patients (110 male, 104 female; age range, 12-77 years) who were routinely referred for cerebral angiography. Middle cerebral artery spasm was graded as mild (</= 25% of vessel caliber reduction) and moderate to severe (> 25% reduction). Angle-corrected blood velocity measurements were obtained using a 2.5-MHz probe. The velocity values were reexpressed as a percentage of the mean of normal reference values for the relevant age, for subjects younger than 40 years, and for sex. RESULTS: The prevalence of spasm among 335 arteries studied was 8.1% for mild and 12.8% for moderate to severe middle cerebral artery narrowing. For distinguishing all or moderate to severe vasospasm from lesser grades of vasospasm, peak systolic velocity was the best parameter. Areas under ROC curves for all and moderate to severe middle cerebral artery spasms were 0.83 and 0.92, respectively. After standardization, the ROC areas increased significantly (p < 0.05) for all, to 0.86, and only slightly, to 0.93, for moderate to severe spasms. For all grades of middle cerebral artery spasm, the best efficiencies were found at standardized velocity value of 170%. CONCLUSION: The accuracy of transcranial color Doppler sonography is high in the identification of middle cerebral artery spasm. Standardization of velocities with respect to age and sex increases the accuracy of the method in diagnosing mild middle cerebral artery spasms.
OBJECTIVE: The overall accuracy of transcranial Doppler sonography in the diagnosis of middle cerebral artery spasm has not been established. Moreover, the factors of age and sex have not been addressed in most studies. In this article, we present receiver operating characteristic (ROC) curve analysis of the accuracy of transcranial color Doppler sonography in diagnosing middle cerebral artery spasm on the basis of flow velocities standardized for age and sex. SUBJECTS AND METHODS. We prospectively studied 214 consecutive patients (110 male, 104 female; age range, 12-77 years) who were routinely referred for cerebral angiography. Middle cerebral artery spasm was graded as mild (</= 25% of vessel caliber reduction) and moderate to severe (> 25% reduction). Angle-corrected blood velocity measurements were obtained using a 2.5-MHz probe. The velocity values were reexpressed as a percentage of the mean of normal reference values for the relevant age, for subjects younger than 40 years, and for sex. RESULTS: The prevalence of spasm among 335 arteries studied was 8.1% for mild and 12.8% for moderate to severe middle cerebral artery narrowing. For distinguishing all or moderate to severe vasospasm from lesser grades of vasospasm, peak systolic velocity was the best parameter. Areas under ROC curves for all and moderate to severe middle cerebral artery spasms were 0.83 and 0.92, respectively. After standardization, the ROC areas increased significantly (p < 0.05) for all, to 0.86, and only slightly, to 0.93, for moderate to severe spasms. For all grades of middle cerebral artery spasm, the best efficiencies were found at standardized velocity value of 170%. CONCLUSION: The accuracy of transcranial color Doppler sonography is high in the identification of middle cerebral artery spasm. Standardization of velocities with respect to age and sex increases the accuracy of the method in diagnosing mild middle cerebral artery spasms.
Authors: J Krejza; W Rudzinski; M A Pawlak; M Tomaszewski; R Ichord; J Kwiatkowski; D Gor; E R Melhem Journal: AJNR Am J Neuroradiol Date: 2007-09 Impact factor: 3.825
Authors: Miroslaw Swiercz; Jan Kochanowicz; John Weigele; Robert Hurst; David S Liebeskind; Zenon Mariak; Elias R Melhem; Jaroslaw Krejza Journal: Neuroinformatics Date: 2008-08-13