[ELECTROPHYSIOLOGIC ANALYSIS OF MENTAL ARITHMETIC TASK BY THE "MINIMUM SPANNING TREE" METHOD]

In the present study basic arithmetic induced rearrangements in functional connections of the brain were investigated by using graph theoretical analysis what becomes increasingly important both in theoretical neuroscience and also in clinical investigations.

During mental arithmetic operations (working) memory plays an important role, but there are only a few studies in which an attempt was made to separate this effect from the process of arithmetic operations themselves. The goal of our study was to separate the neural networks involved in cognitive functions.

As an attempt to clarify this issue the graph-theoretical "minimal spanning tree" method was used for the analysis of EEG recorded during task performance. The effects of passive viewing, number recognition and mental arithmetic on PLI based minimal spanning trees (MST) were investigated on the EEG in young adults (adding task: 17 subjects; passive viewing and number recognition: 16 subjects) in the θ (4-8 Hz) frequency band.

Occipital task relevant synchronization was found by using the different methods, probably related to the effect of visual stimulation. With respect to diameter, eccentricity and fraction of leafs different task-related changes were found.

It was shown that the task related changes of various graph indices are capable to identify networks behind the various relevant dominant functions. Thus the "minimal spanning tree" method is suitable for the analysis of the reorganization of the brain with respect to cognitive functions.