MRI Protocol and Imaging Analysis
At our institution, an acute stroke MRI protocol has been used routinely since March 2001 to evaluate the condition of patients with suspected ischemic stroke. T2*-weighted gradient-echo imaging has been part of the protocol since September 2003. From March 2001 to August 2003, instead of T2*-weighted imaging, unenhanced CT was frequently performed to exclude ICH, and T2*-weighted imaging was occasionally used in the acute stroke MRI protocol. Among 279 patients, those who did not undergo T2*-weighted imaging were excluded from the study population. The acute stroke MRI protocol, which consisted of T2*-weighted gradient-echo imaging, diffusion-weighted imaging, 3D time-of-flight MR angiography, FLAIR imaging, perfusion-weighted imaging, and contrast-enhanced MR angiography, was performed in a single session within 20 min. Patients with unstable vital signs or contraindications to MRI were excluded. The MRI parameters were as follows: T2*-weighted gradient-echo sequence (TR/TE, 400/30; flip angle, 20°; thickness, 5 mm; number of axial slices, 20; intersection gap, 2 mm; matrix, 256 × 192; number of excitations, 2; field of view, 250 mm; acquisition time, 2 min 32 sec); diffusion-weighted spin-echo echo-planar image (TR/TE, 7,500/84; thickness, 5 mm; number of axial slices, 20; matrix, 128 × 128; field of view, 250 mm; number of excitations, 1; acquisition time, 1 min; b values, 0 and 2,000 sec/mm2); 3D time-of-flight MR angiogram (TR/TE, 25/2; flip angle, 20°; matrix, 512 × 512; field of view, 250 mm; number of excitations, 1; acquisition time, 3 min 56 sec); FLAIR image (TR/TE, 10,002/97; number of axial slices, 20; thickness, 5 mm; intersection gap, 2 mm; matrix, 256 × 192; field of view, 250 mm; acquisition time, 2 min 20 sec); perfusion-weighted MR image (TR/TE, 2,000/40; flip angle, 90°; number of axial slices, 10; thickness, 5 mm; intersection gap, 2 mm; matrix, 512 × 512; field of view, 250 mm; number of excitations, 1; acquisition time, 1 min 22 sec); 3D contrast-enhanced MR angiogram (TR/TE, 6/1; flip angle, 20°; matrix, 512 × 512; field of view, 250 mm; number of excitations, 1; acquisition time, 46 sec). The scanning time for the entire MRI protocol was 11 min 56 sec. This MRI protocol was performed between 8:00 am and 10:00 pm. Between 10:00 pm and 8:00 am, patients with suspected acute ischemic stroke were examined with a CT protocol that included unenhanced CT, perfusion CT, and CT angiography.
Microbleeds were defined as homogeneous rounded areas of signal loss less than 5 mm in diameter without surrounding edema on T2*
-weighted images. Symmetric signal loss in the globus pallidus, flow voids, and large ICH foci were excluded. The number and location of microbleeds were assessed on initial T2*
-weighted gradient-echo images. Microbleeds were classified as absent (grade 1), mild (grade 2; total number of microbleeds, 1-2), moderate (grade 3; total number of microbleeds, 3-10), and severe (grade 4; total number of microbleeds, > 10) according to a grading scale described previously [5
] (Figs. 1A
, and 1D
). Follow-up T2- and T2*
-weighted images were obtained 1-3 days after thrombolysis. The hemorrhage was graded as symptomatic hemorrhage if any neurologic deterioration had occurred within the first 48 hr that could be attributed to the presence of such hemorrhage. According to the number and location of microbleeds, we evaluated the prevalence and severity of subsequent hemorrhagic transformation on follow-up MR images. Imaging findings were reviewed by two neuroradiologists without knowledge of clinical information or treatment assignment. Their consensus determined the MRI findings.
Multivariate logistic regression analysis was used to assess the relation between microbleeds and risk factors for stroke, including age, sex, hypertension, diabetes, hyperlipidemia, smoking, and atrial fibrillation, and the relation between subsequent hemorrhagic transformation and age, sex, baseline NIH-SS score, hypertension, diabetes, hyperlipidemia, previous stroke, atrial fibrillation, and microbleeds. Fisher's exact test was used to assess the relation between number of microbleeds and severity of hemorrhagic transformation.