A. Suhhova, M. Bachmann, J. Lass, V. Tuulik and H. Hinrikus Department of Biomedical Engineering, Technomedicum of Tallinn University of Technology, Tallinn, Estonia Abstract— This study is aimed to test the EEG spectral asymmetry index (SASI) in different brain states. The SASI parameter was previously developed for depres- sive disorder. In this study it was assumed that micro- wave exposure can alter the brain state and the SASI could estimate the changes. For this, 15 healthy subjects were exposed to 450 MHz microwave radiation modu- lated at 40 Hz and 70 Hz frequencies. The field power density at the scalp was 0.16 mW/cm 2 . The analysis re- vealed significant rise in SASI value caused by 40 Hz modulated microwave for the whole group. Microwave modulated at 70 Hz also raised the SASI level, however the increase did not reach the level of significance. De- spite the large increase in average SASI level, it still remained negative while in case of depressive disorder the SASI level was positive. Experimental results con- firm that the EEG SASI detects changes in the brain state caused by different reasons. Keywords— EEG, spectral asymmetry index (SASI), micro- wave, depressive disorder I. INTRODUCTION During last decades understanding of the brain activity became a topic of major interest. Nowadays the modern methods of brain imaging are applied to get objective in- formation about changes in brain physiology. EEG is easily available and incredibly cost-effective method. For instance, some steps toward understanding and assessment of mental fatigue and dementia [1, 2] have been reported. The electroencephalogram (EEG) analysis provid- ed promising results for evaluation of Alzheimer’s disease [3 – 5]. Nevertheless, EEG as a technique to evaluate brain dis- orders is often left aside. Alterations in brain activity related to different brain states definitely appear in the EEG. How- ever, those incidents are difficult to detect. Excellent signal processing algorithms are needed for EEG interpretation. Several measures based on the EEG analysis have been successfully employed for evaluation of the brain physio- logical state and different diseases [6 – 9]. The EEG spec- tral mean frequency was used to provide computational sleep depth analysis [6]. The mean frequency represents the center of gravity of the EEG amplitude spectrum 0.58–30.1 Hz [6]. An epileptic abnormality index was constructed from the EEG frequency band power and power inter- hemispheric asymmetry parameters comparing powers of the low (delta and theta) and high (alpha and beta) frequen- cy bands between the brain hemispheres [7]. The revised extended brain symmetry index has been proposed to detect interhemispheric asymmetry, in particular with an emphasis on the detection of cerebral ischemia [8]. However, there is still a wide range of mental disorders in case limited data is available. For example the diagnosis of depression is tradi- tionally based on evaluation of the intensity of subjective and clinical symptoms by psychiatrists (M.I.N.I. interview, Hamilton Depression Rating Scale). Nevertheless, subjec- tive symptoms of depression are accompanied by objective alterations in the brain bioelectrical activity and in the EEG signal. Therefore, the indicators based on the EEG analysis can be fruitful for diagnostic purpose. In this work we study a method, which has previously shown effective for detection of depressive disorder [10]. It is based on analysis of the EEG frequency spectrum, where- as balance of different spectral band powers can be evaluat- ed in an arbitrary EEG channel and hemisphere. This meth- od presumes that the EEG beta band includes useful information for evaluation of depression, whereas the EEG theta band is stable and not affected by a disease. A spectral asymmetry index (SASI) is calculated as a relative differ- ences in power of two EEG special frequency bands select- ed higher and lower of the EEG spectrum maximum. The EEG central frequency band round the spectrum maximum (alpha band) is excluded from the calculations. The bounda- ry frequencies of the bands are specially selected for each individual and related to the frequency of the EEG spectrum maximum. The frequency bands selected for calculation are close to the traditional beta and theta bands, but can be shifted in the case of low or high alpha frequency. The effi- ciency of the previously developed method is studied under the exposure of microwave radiation presuming that the exposure influences balance of the EEG frequencies. Á. Jobbágy (Ed.): 5th European IFMBE Conference, IFMBE Proceedings 37, pp. 1206–1209, 2011. www.springerlink.com EEG Spectral Asymmetry Index Reveals Effect of Microwave Radiation