O3‐13‐05: Evaluation of Event Related Potentials and Quantitative Eeg as Useful Biomarkers for the Diagnosis of Early‐stage Alzheimer's Disease

This is the authors' abstract. We don't add key points for this paper.

Given the limitations of psychometric testing, event-related potentials (ERPs) and quantitative EEG (qEEG) have emerged as potentially useful cognitive biomarkers for Alzheimer's disease (AD). However, lack of standardization in data acquisition and analysis methods have limited the adoption of these techniques in clinical environments on real patients. In our study, we investigated whether ERPs and qEEG collected with standardized methods can be useful biomarkers for detection of early-stage AD in outpatient settings. 100 subjects between 60 and 90 years old with probable mild AD (MMSE 21-26) and 100 age-matched Controls (MMSE 27-30) were recruited at seven clinical sites in US. All study subjects were administered psychometric and ERP/EEG tests. Subjects with probable AD also had blood testing, an MRI scan, and provided an optional CSF sample. The ERP/EEG protocol for the study consisted of a two-deviant auditory oddball paradigm followed by 3min of resting EEG. The test was administered in outpatient settings, using an integrated hardware/software system for data collection and analysis.The ERP and EEG data collected are currently been used to train an ensemble of classifiers to discriminate subjects with mild AD from cognitively normal individuals. Correlations of ERP and qEEG data with psychometric measures, MRI volumetry and ApoE genetic profile are also been analyzed. Preliminary classification results show an accuracy of 82%, sensitivity of 77%, specificity of 86%, and an area under the ROC curve of 0.824. Comparison of test results for ERP and EEG tests administered at the different clinical sites show little inter-site variability. Final classification results, correlation of ERP/EEG data to standardized psychometric tests and hippocampal volume, and comparison of EEG/ERP data in ApoE4 positive (ApoE4+) and (ApoE4-) AD subjects will be available at the time of the conference. The use of an integrated hardware/software system for ERP/EEG testing and analysis can address the main limitations that have hindered the adoption of neuro-electrophysiological measures as useful biomarkers for AD in clinical settings. Preliminary results from the study support the classification approach to the analysis of data from ERP and EEG tests as an effective tool for early diagnosis of AD.