[Retrospective analysis of routine-MRI and correlation with aetiology of dementia, severity and neuropsychology in a memory-clinic population].

Abstract

Background: The contribution of potential treatable dementia aetiologies diagnosed using cerebral imaging varied considerably in previous studies and was not evaluated in a recent larger German sample of patients from a memory clinic.

Material and methods: MRI images of 502 patients were retrospectively reassessed. Beside the proportion of potentially treatable dementia aetiology, the extent of whole brain atrophy (semiquantitative) and vascular white matter lesions were assessed.

Results: Mean age of the patients was 63.7 ± 13.1 years; 49 % were female, mean MMST was 24.2 ± 5.5. In 74 % there was an agreement between the clinical dementia syndrome and MRI. 9 % (45 patients) had clearly discrepant imaging results, according to MRI criteria (21 × ischaemia, 20 × normal pressure hydrocephalus (NPH), 4 × intracerebral haemorrhage). These patients could not not be differentiated using age and MMST alone as clinical criteria. There was a significant correlation between global brain atrophy and MMST (r = -0.32; p < 0.001) and white matter lesion score (r = 0.28; p < 0.001).

Conclusion: In 9 % there was a clear discrepancy between MRI results and the clinical syndrome diagnosis in memory-clinic patients. As known from earlier studies and current German 3 rd generation guidelines, it is reasonable to perform MRI imaging in dementia to improve the aetiological and differential diagnoses and to detect a different aetiology that can be missed using the clinical dementia criteria alone.