Topographic classification of EEG patterns in Huntington's disease.

The aim of this study is to perform a topographic classification of electroencephalographic (EEG) patterns in subjects affected by the Huntington's disease (HD). The alpha activity is a discriminating feature for HD, as its amplitude reduction turns out to be a clear mark of the illness. When used as input variable to a supervised neural network, a good classification of pathological patterns and control ones is achieved with high values of sensitivity and specificity. It should be useful to get more insight into the local discriminating capabilities of the alpha rhythm by implementing a neural network approach to classify EEG patterns extracted from groups of channels corresponding to specific regions of the scalp. Receiver operating characteristic (ROC) curve analysis enables one to label each region with the value of the area under the curve, thus providing a local significance for HD classification. A reduction of the area when processing regions of the scalp, with respect to the whole, suggests that all channels provide significant contribution to HD pattern discrimination. These results can be interpreted as an effect of an abnormal subcortical modulation of the alpha rhythm, due to the dysfunctional action of the thalamus on the cortical activities. In a further study, morphometric features of thalamus and basal ganglia, evaluated by MRI, will be matched with the electrophysiological findings.