OBJECTIVE: To determine the relative effect of carotid plaque heterogeneity and echogenicity as measured by computer on the incidence of ipsilateral cerebral infarction and symptoms of cerebrovascular disease. MATERIALS AND METHODS: 138 patients with 209 carotid plaques producing > 50% stenosis were studied. 110 plaques were from symptomatic sides. All patients had computer tomography (CT) brain scans. Images of the carotid plaques obtained by duplex scanner were transferred to a computer. Using an image analysis program plaque echogenicity, measured as the grey scale median (GSM) and heterogeneity, evaluated as the heterogeneity index (HI) was calculated. RESULTS: 42% of plaques were associated with cerebral infarction. Symptomatic plaques and those associated with cerebral infarction were more echolucent and less heterogeneous than asymptomatic plaques and those not associated with cerebral infarction. Plaques with GSMs below or equal to 32 were associated with a significantly higher incidence of cerebral infarction. Plaques with GSMs below or equal to 32 were associated with a significantly higher incidence of cerebral infarction and symptoms of cerebrovascular disease than those with GSMs above this level (p < 0.01). Plaques with HIs below or equal to 20 were associated with a significantly higher incidence of symptoms of cerebrovascular disease as compared to those with HIs above this level (p < 0.01). CONCLUSIONS: Computer analysis of carotid plaque morphology is an objective tool that can identify high risk plaques. The potential of such analysis in the identification of asymptomatic patients with advanced carotid stenosis at a high risk of stroke should be explored in a natural history study.
OBJECTIVE: To determine the relative effect of carotid plaque heterogeneity and echogenicity as measured by computer on the incidence of ipsilateral cerebral infarction and symptoms of cerebrovascular disease. MATERIALS AND METHODS: 138 patients with 209 carotid plaques producing > 50% stenosis were studied. 110 plaques were from symptomatic sides. All patients had computer tomography (CT) brain scans. Images of the carotid plaques obtained by duplex scanner were transferred to a computer. Using an image analysis program plaque echogenicity, measured as the grey scale median (GSM) and heterogeneity, evaluated as the heterogeneity index (HI) was calculated. RESULTS: 42% of plaques were associated with cerebral infarction. Symptomatic plaques and those associated with cerebral infarction were more echolucent and less heterogeneous than asymptomatic plaques and those not associated with cerebral infarction. Plaques with GSMs below or equal to 32 were associated with a significantly higher incidence of cerebral infarction. Plaques with GSMs below or equal to 32 were associated with a significantly higher incidence of cerebral infarction and symptoms of cerebrovascular disease than those with GSMs above this level (p < 0.01). Plaques with HIs below or equal to 20 were associated with a significantly higher incidence of symptoms of cerebrovascular disease as compared to those with HIs above this level (p < 0.01). CONCLUSIONS: Computer analysis of carotid plaque morphology is an objective tool that can identify high risk plaques. The potential of such analysis in the identification of asymptomatic patients with advanced carotid stenosis at a high risk of stroke should be explored in a natural history study.
Authors: J J Connors; D Seidenwurm; J C Wojak; R W Hurst; M E Jensen; R Wallace; T Tomsick; J Barr; C Kerber; E Russell; G M Nesbit; A J Fox; F Y Tsai Journal: AJNR Am J Neuroradiol Date: 2000-03 Impact factor: 3.825