Key points are not available for this paper at this time.
BACKGROUND: In view of the increasing recognition of cerebral microbleeds (MCBs) with MRI, there is a need to validate their detection in post-mortem brains in patients with cerebrovascular diseases and dementia. MATERIALS AND METHODS: Out of 20 post-mortem brains of patients with Alzheimer dementia and with different cerebrovascular lesions, 45 large sections of the cerebral hemispheres, brainstem and cerebellum were submitted to a 7.0-T T₂*-weighted MRI, and afterwards compared to the histological detection of haematomas, MCBs and mini-bleeds (MNBs). RESULTS: The sensitivity, specificity, predictive positive value and predictive negative value of the T₂* imaging to detect MCBs and MNBs were excellent for those in the cortico-subcortical regions. There was a significant overestimation of MNBs in the striatum due to iron deposits unrelated to old haemorrhages. Also in the deep white matter, 42% of MNBs were not detected, while 31% of T₂* hyposignals were not due to MNBs but to vessels filled with post-mortem thrombi. CONCLUSIONS: When evaluating the 'bleeding load' with 7.0-T T₂*-weighted MRI in post-mortem brain sections of patients with dementia and vascular risk factors, only quantification of small cerebral bleeds in the cortico-subcortical regions is reliable.
Reuck et al. (Sat,) studied this question.