Analysing specificity of a bipolar EEG measurement

The objective in bioelectric measurements such as ECG and EEG is to register the signal arising from sources in the region of interest. It is also desired that signal-to-noise ratio (SNR) of a measurement is high. The sensitivity of an ideal measurement should focus on and be greater on the target areas in comparison to other areas of the volume conductor. Previously the half-sensitivity volume (HSV) has been applied to describe how focused the measurement is. In this paper we introduce a concept of the half-sensitivity ratio (HSR) which describes how well the sensitivity is concentrated in HSV compared to other source regions i.e. how specific the measurement is to the sources in HSV. Further we may have different region of interests (ROI) to which the measurements are wanted to be specific. Then the concept is called region of interest sensitivity ratio (ROISR). We present here an application of the HSR in analysing sensitivity distributions of bioelectric measurements. We studied the effects of interelectrode distance and the scalp/skull/brain resistivity ratio on the HSR of a bipolar EEG measurement with a three-layer spherical head model. The results indicate that when the focus of interest is on cortical activity more specified and concentrated sensitivity distributions are achieved with smaller interelectrode distances. Further a preliminary measurement with visual evoked potentials provides evidence of the relationship between HSR and SNR of a measurement.