Standard resting state electroencephalographic power density computed from posterior bipolar channel characterizes patients with Alzheimer’s disease

Abstract

AbstractBackgroundAbnormalities in cortical sources of resting‐state eyes closed electroencephalographic (rsEEG) rhythms recorded by hospital settings (10‐20 montage) with 19 scalp electrodes characterized Alzheimer’s disease (AD) from preclinical to dementia stages. An intriguing rsEEG application is the monitoring and evaluating of AD progression in large populations with few electrodes in low‐cost devices. Here we assessed whether the above‐mentioned abnormalities could be observed from fewer scalp electrodes in patients with mild cognitive impairment (ADMCI) or dementia (ADD) due to AD.MethodClinical and rsEEG data acquired in hospital settings (10‐20 montage) from 88 ADD and 75 ADMCI participants and 70 age‐, education‐, and sex‐matched normal elderly controls (Nold) were available in an Italian‐Turkish archive (PDWAVES Consortium; www.pdwaves.eu). Standard spectral Fast Fourier Transform (FFT) analysis of rsEEG data for individual delta, theta, and alpha frequency bands was computed from 6 monopolar scalp electrodes to derive bipolar C3‐P3, C4‐P4, P3‐O1, and P4‐O2 markers.ResultThe ADMCI and ADD groups showed increased delta/alpha and theta/alpha power density at the C3‐P3, C4‐P4, P3‐O1, and P4‐O2 bipolar channels compared to the Nold group (Figure 1). The best classification accuracy (area under the receiver operating characteristic curve) between the ADD and ADMCI vs Nold individuals was reached using alpha2/theta power density computed at the C3‐P3 bipolar channel (AUROC for ADMCI vs Nold = 81%; AUROC for ADD vs Nold = 90%; Figure 2).ConclusionStandard rsEEG power density computed from six posterior bipolar channels characterized AD status from preclinical to dementia stages. These results may pave the way toward diffuse clinical applications in the health monitoring of dementia using low‐cost EEG systems with a strict number of electrodes in lower‐ and middle‐income countries.