Cerebral arterial infarction and diffusion tensor imagingPurpose. Diffusion-weighted imaging (3 directions) and diffusion tensor imaging (9 directions) were compared for their sensitivity to detect ischemic lesion.Materials and methods. 41 patients (18 supposed transient ischemic attacks, 23 arterial stroke, MRI <= 48h) presenting with stroke syndrome were imaged using two diffusion sequences (3 directions, 2 excitations, acquisition time: 40 seconds and 9 directions, 1 excitation, acquisition time: 50 seconds, other parameters being identical). The following variables were compared (Wilcoxon test): number and extent of lesions, MRI signal and absolute ADC values, relative signal intensity and ADC ratio. Fractional anisotropy was measured within the principal lesion on the 9 directions sequence.Results. 24 patients presented a lesion on both sequences. Four of them (including one TIA) had more lesions (<5 mm) on the 9 directions sequence than on the 3 directions sequence. Both sequences were normal for the 17 remaining patients (TIA=4, differential diagnosis=7, indeterminate=6). No significant difference was found for the other study parameters between 3 vs 9 directions sequences. There was a 10% decrease of the mean fractional anisotropy in ischemic lesion.Conclusion: By showing more lesions than standard diffusion-weighted imaging and providing quantitative anisotropy measurements, diffusion tensor imaging could replace the standard 3 directions diffusion-weighted sequence.