[Use of artificial neural networks in recognition of types of thinking by EEG]

Reflection in EEG parameters of the process of solving two types of mental tasks, i.e. spatial-imaginary and verbal-logical, was investigated in nine test-subjects. Mental efforts were found to specify EEG spatiofrequency patterns depending on the task type and subject's identity. These patterns were unique and reproducible enough to enable identification of this or the other type of mental task within several seconds of EEG record. Recognition was executed with the use of an artificial neural network algorithm in order to simultaneously weigh a wealth of signs of a specific EEG pattern. Recognition precision was 85% on the average.