This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

No abstract

Background: There is still a need for simple, noninvasive, and inexpensive methods to diagnose the causes of cognitive impairment and dementia. In this study, contemporary statistical methods were used to classify the clinical cases of cognitive impairment based on electroencephalograms (EEG). Methods: An EEG database was established from seven different groups of subjects with cognitive impairment and dementia as well as healthy controls. A classifier was created for each possible pair of groups using statistical pattern recognition (SPR). Results: A good-to-excellent separation was found when differentiating cases of degenerative disorders from controls, vascular disorders, and depression but this was less so when the likelihood of comorbidity was high. Conclusions: Using EEG with SPR seems to be a reliable method for diagnosing the causes of cognitive impairment and dementia, but comorbidity must betaken into account.

No abstract

Differential diagnosis of Alzheimer's disease (AD) and dementia with Lewy bodies (DLB) remains challenging; currently the best discriminator is striatal dopaminergic imaging. However this modality fails to identify 15–20% of DLB cases and thus other biomarkers may be useful. It is recognised electroencephalography (EEG) slowing and relative medial temporal lobe preservation are supportive features of DLB, although individually they lack diagnostic accuracy. Therefore, we investigated whether combined EEG and MRI indices could assist in the differential diagnosis of AD and DLB.Seventy two participants (21 Controls, 30 AD, 21 DLB) underwent resting EEG and 3 T MR imaging. Six EEG classifiers previously generated using support vector machine algorithms were applied to the present dataset. MRI index was derived from medial temporal atrophy (MTA) ratings. Logistic regression analysis identified EEG predictors of AD and DLB. A combined EEG-MRI model was then generated to examine whether there was an improvement in classification compared to individual modalities.For EEG, two classifiers predicted AD and DLB (model: χ2 = 22.1, df = 2, p < 0.001, Nagelkerke R2 = 0.47, classification = 77% (AD 87%, DLB 62%)). For MRI, MTA also predicted AD and DLB (model: χ2 = 6.5, df = 1, p = 0.01, Nagelkerke R2 = 0.16, classification = 67% (77% AD, 52% DLB). However, a combined EEG-MRI model showed greater prediction in AD and DLB (model: χ2 = 31.1, df = 3, p < 0.001, Nagelkerke R2 = 0.62, classification = 90% (93% AD, 86% DLB)).While suggestive and requiring validation, diagnostic performance could be improved by combining EEG and MRI, and may represent an alternative to dopaminergic imaging.