Search

Diffusion-weighted imaging (DWI) in children with diarrhoea associated haemolytic uraemic syndrome (D+HUS) and cerebral involvement was evaluated retrospectively.DWI within 24 h of onset of neurological symptoms. The apparent diffusion coefficient (ADC) was measured in grey/white matter and correlated with clinical and laboratory findings.DWI was abnormal in all. Abnormal ADC was detected in the supratentorial white matter (6/12) and cortex (1/12), the basal ganglia (5/12), the thalami (4/12), and the cerebellum (1/12). ADC was reduced in 5/12, increased in 4/12, and both in 3/12. Mean serum sodium was lower in patients with DWI abnormalities affecting the white matter (6/12), than in those with basal ganglia/thalamic involvement (6/12). Neurological outcome was normal in 4/11 and abnormal in 7/11, and 1 patient died, outcome did not correlate to either localisation or type of DWI abnormality.In D+HUS with neurological symptoms, early DWI may reveal abnormal ADC not only in the basal ganglia/thalami, but also in the white matter/cortex. Besides thrombotic microangiopathy, toxic effects of shiga toxin, azotaemia and hyponatraemia / hypoosmolality may be involved in cerebral involvement in children with D+HUS. Findings on early MRI seem not to predict clinical course or outcome.• DWI MR imaging may detect early CNS involvement in haemolytic uraemic syndrome • Different pathogenetical mechanisms may contribute to the CNS disease in HUS • Early MRI findings do not seem to allow prediction of clinical outcome.

Twelve patients with ICA stenosis underwent dynamic perfusion computed tomography (CT) and positron emission tomography (PET) studies at rest and after acetazolamide challenge. Cerebral blood flow (CBF) maps on perfusion CT resulted from a deconvolution of parenchymal time-concentration curves by an arterial input function (AIF) in the anterior cerebral artery as well as in both anterior choroidal arteries. CBF was measured by [(15)O]H(2)O PET using multilinear least-squares minimization procedure based on the one-compartment model. In corresponding transaxial PET scans, CBF values were extracted using standardized ROIs. The baseline perfusion CT-CBF values were lower in perfusion CT than in PET (P0.05). CBF values obtained by perfusion CT were significantly correlated with those measured by PET before (P0.05) and after (P0.01) acetazolamide challenge. Nevertheless, the cerebrovascular reserve capacity was overestimated (P=0.05) using perfusion CT measurements. The AIF selection relative to the side of carotid stenosis did not significantly affect calculated perfusion CT-CBF values. In conclusion, the perfusion CT-CBF measurements correlate significantly with the PET-CBF measurements in chronic carotid stenotic disease and contribute useful information to the evaluation of the altered cerebral hemodynamics.