Tung G A, Julius B D, Rogg J M
Department of Diagnostic Imaging, Brown Medical School and Rhode Island Hospital, 593 Eddy Street, Providence, Rhode Island 02903, USA.
Neuroradiology. 2003 Jun;45(6):357-62. doi: 10.1007/s00234-003-0994-0. Epub 2003 May 8.
We evaluated a method for quantifying vasogenic edema (VE) on MRI to identify brain hematomas caused by neoplasms. We performed a blinded review of 68 acute and subacute hematomas caused by neoplasms (22), hypertension or presumed amyloid angiopathy (27), or vascular malformations (19). The extent of vasogenic edema was quantified on an axial T2-weighted image using the VE ratio: the maximum width of high signal surrounding a hematoma and the mean diameter of the hematoma. Hematomas caused by neoplasm were associated with more vasogenic edema (mean VE ratio 104%+/-15%; mean VE width 2.4+/-0.7 cm) than non-neoplastic hematomas (mean VE ratio 37%+/-5%; P =0.001). When the width of VE was equal to or more half the diameter the hematoma (VE ratio 50%), the positive predictive value for tumor was 66%; when it was equal to or more than the diameter, the positive predictive value was 71%. All six hematomas with VE ratios > or = 150% were caused by neoplasm.
我们评估了一种在磁共振成像(MRI)上量化血管源性水肿(VE)的方法,以识别由肿瘤引起的脑血肿。我们对68例由肿瘤(22例)、高血压或疑似淀粉样血管病(27例)或血管畸形(19例)引起的急性和亚急性血肿进行了盲法评估。使用VE比率在轴向T2加权图像上对血管源性水肿的程度进行量化:血肿周围高信号的最大宽度与血肿的平均直径。与非肿瘤性血肿(平均VE比率37%±5%;P = 0.001)相比,肿瘤引起的血肿伴有更多的血管源性水肿(平均VE比率104%±15%;平均VE宽度2.4±0.7 cm)。当VE宽度等于或大于血肿直径的一半(VE比率50%)时,肿瘤的阳性预测值为66%;当它等于或大于直径时,阳性预测值为71%。所有6例VE比率≥150%的血肿均由肿瘤引起。