SUMMARY: Different MR imaging patterns of cerebral fat embolism have been reported in the literature without a systematic review. Our goal was to describe the patterns, explore the relationship between disease course and the imaging patterns, and discuss the underlying mechanism. We reveal 5 distinctive MR imaging patterns: 1) scattered embolic ischemia occurring dominantly at the acute stage; 2) confluent symmetric cytotoxic edema located at the cerebral white matter, which mainly occurs at the subacute stage; 3) vasogenic edematous lesions also occurring at the subacute stage; 4) petechial hemorrhage, which persists from the acute to the chronic stage; and 5) chronic sequelae, occurring at late stage, including cerebral atrophy, demyelinating change, and sequelae of infarction or necrosis. Underlying mechanisms of these imaging patterns are further discussed. Recognition of the 5 evolving MR imaging patterns of cerebral fat embolism may result in adjustment of the appropriate management and improve the outcome. ABBREVIATION: CFE ϭ cerebral fat embolism

Background/Aims: To investigate whether diffusion tensor imaging (DTI) is more sensitive than conventional MRI at detecting cognitive deterioration in patients with subcortical ischemic vascular disease (SIVD). Methods: Forty-two SIVD patients had a diagnosis of no cognitive impairment (NCI), vascular cognitive impairment/no dementia or vascular dementia (VaD). Whole-brain DTI histography and routine MRI were performed on these participants. Results: There were significant differences between cognitively impaired patients and NCI subjects in mean diffusivity and fractional anisotropy in either whole-brain white matter (WBWM) or in normal-appearing white matter (NAWM). All DTI indices within either WBWM or NAWM were found to be significantly correlated with both the attention-executive and memory measures in SIVD subjects. Lacune numbers and T₂-weighted lesions correlated only with attention-executive measures, whereas hippocampal volumes correlated only weakly with memory measures. Whole-brain gray matter volumes correlated with Z scores for all cognitive domains but language. After VaD patients had been excluded from the analysis, cognitive measures remained significantly correlated with some of the DTI indices, but not with conventional MRI findings. Conclusions: Compared with conventional MRI, whole-brain DTI is a more reliable and sensitive technique for the early detection of cognitive impairment in SIVD patients.