PURPOSE: Diffusion-weighted MRI (DWI) is well known to be sensitive in the detection of hyperacute infarct, but has not been systematically investigated in patients with acute or subacute infarct. We evaluated the usefulness of diffusion-weighted MRI in assessing the various stages of brain infarct. MATERIALS AND METHODS: Fifty-five consecutive patients with symptoms of brain infarct underwent fast spinecho T2-weighted MRI (T2W1) and DWI. Using only a brief clinical history, two radiologists first attemptelto detect the lesion using T2W1, which was then compared with DWI. The usefulness of the latter was then evaluated in terms of the following criteria: 1) Its abilility to detect a lesion not seen at T2WI (detection); 2) localization of the responsible ischemic focus among multiple high-signal intensities seen at T2WI (localization); 3) conspicuity of a lesion which was subtle at T2WI (conspicuity); 4) detection of multiple lesions (multiplicity). RESULTS: DWI was useful in 44 of 55 patients (80%), including 9 of 9 (100%) with hyperacute infarct (<6 hours), 20 of 27 (74%) with acute infarct (<48 hours), and 15 of 19 (79%) with subacute infarct (<2 weeks). Among the nine patients at the hyperacute stage, DWI was useful for detection of the lesion in six (67%), for localization, 4 (44%) in one (11%), for conspicuity in four (44%), and for multiplicity in five (56%); at the acute stage (20 patients), for detection in three (15%), for localization in ten (50%), for conspicuity in eight (40%), and for moltiplicity in five (25%); and at the subacute stage (15 patients), for detection in three (20%), for localization in nine (60%), for conspicuity in two (13%), and for multiplicity in three (20%). CONCLUSION: DWI is very sensitive for the diagnosis of hyperacute infarct. In the assessment of this, it is useful during the acute or subacute period for the detection of small lesions, the localization of ongoing lesions among multiple high signal intensities seen at T2WI, and the determination of lesion conspicuity.
Brain ; Diagnosis ; Humans ; Magnetic Resonance Imaging* ; Stroke*