Using clinical information and nonlinear EEG analysis for diagnosis of dementia

The analysis of electrocardiograms presents a particularly difficult problem due to a number of factors, including the lack of specificity of the signal and the inability to map the scalp potential to physiological parameters. In the work described here, nonlinear EEG analysis based on computation of cortical potential followed by nonlinear analysis based on the computation of degree of variability is combined with imaging results and clinical parameters to form a diagnostic model for dementia diagnosis. These parameters are combined through the use of an intelligent agent model that uses a knowledge-based system and a neural network model in addition to the biomedical signal analyzer.