EEG-BASED, NEURAL-NET PREDICTIVE CLASSIFICATION OF ALZHEIMERS-DISEASEVERSUS CONTROL SUBJECTS IS AUGMENTED BY NONLINEAR EEG MEASURES

Attempts to classify Alzheimer's disease (AD) subjects versus controls using spectral-band measures of electroencephalographic (EEG) data ty pically achieve around 80% success. This study assessed the ability of adding non-linear EEG measures and using a neural-net classification procedure to improve this performance level. The non-linear EEG measur es were estimated correlation dimension (''dimensional complexity,'' o r DCx) and saturation (degree of leveling-off of DCx with increasing e mbedding dimension). In a sample of 39 subjects (14 ADs, 25 controls), it was found that (a) the addition of non-linear EEG measures improve d the classification accuracy of the AD/control status of subjects, an d (b) a back-percolation neural net predictively classified the subjec ts much better than the standard linear techniques of multivariate dis criminant analysis or nearest-neighbor discriminant analysis.