Cortical potential imaging using L-curve and GCV method to choose the regularisation parameter.

BACKGROUND: The electroencephalography (EEG) is an attractive and a simple technique to measure the brain activity. It is attractive due its excellent temporal resolution and simple due to its non-invasiveness and sensor design. However, the spatial resolution of EEG is reduced due to the low conducting skull. In this paper, we compute the potential distribution over the closed surface covering the brain (cortex) from the EEG scalp potential. We compare two methods - L-curve and generalised cross validation (GCV) used to obtain the regularisation parameter and also investigate the feasibility in applying such techniques to N170 component of the visually evoked potential (VEP) data. METHODS: Using the image data set of the visible human man (VHM), a finite difference method (FDM) model of the head was constructed. The EEG dataset (256-channel) used was the N170 component of the VEP. A forward transfer matrix relating the cortical potential to the scalp potential was obtained. Using Tikhonov regularisation, the potential distribution over the cortex was obtained. RESULTS: The cortical potential distribution for three subjects was solved using both L-curve and GCV method. A total of 18 cortical potential distributions were obtained (3 subjects with three stimuli each - fearful face, neutral face, control objects). CONCLUSIONS: The GCV method is a more robust method compared to L-curve to find the optimal regularisation parameter. Cortical potential imaging is a reliable method to obtain the potential distribution over cortex for VEP data.

EEG/MEG source imaging: methods, challenges, and open issues.

We present the four key areas of research-preprocessing, the volume conductor, the forward problem, and the inverse problem-that affect the performance of EEG and MEG source imaging. In each key area we identify prominent approaches and methodologies that have open issues warranting further investigation within the community, challenges associated with certain techniques, and algorithms necessitating clarification of their implications. More than providing definitive answers we aim to identify important open issues in the quest of source localization.

Improving the SNR of EEG generated by deep sources with weighted multielectrode leads.

We have developed a multielectrode lead technique to improve the signal-to-noise ratio (SNR) of scalp-recorded electroencephalography (EEG) signals generated deep in the brain. The basis of the method lies in optimization of the measurement sensitivity distribution of the multielectrode lead. We claim that two factors improve the SNR in a multielectrode lead: (1) the sensitivity distribution of a multielectrode lead is more specific in measuring signals generated deep in the brain and (2) spatial averaging of noise occurs when several electrodes are applied in the synthesis of a multielectrode lead. We showed theoretically that within a three-layer spherical head model the sensitivity distributions of multielectrode leads are more specific for deep sources than those of traditional bipolar leads. We also estimated with simulations and with preliminary measurements the total improvement in SNR achieved by both the more specific lead field and spatial averaging. Results obtained with simulations and with experimental measurements show an apparent improvement in SNR obtained with multielectrode leads. This encourages for future development of the method.

New method for analysing sensitivity distributions of electroencephalography measurements.

In this paper, we introduce a new modelling related parameter called region of interest sensitivity ratio (ROISR), which describes how well the sensitivity of an electroencephalography (EEG) measurement is concentrated within the region of interest (ROI), i.e. how specific the measurement is to the sources in ROI. We demonstrate the use of the concept by analysing the sensitivity distributions of bipolar EEG measurement. We studied the effects of interelectrode distance of a bipolar EEG lead on the ROISR with cortical and non-cortical ROIs. The sensitivity distributions of EEG leads were calculated analytically by applying a three-layer spherical head model. We suggest that the developed parameter has correlation to the signal-to-noise ratio (SNR) of a measurement, and thus we studied the correlation between ROISR and SNR with 254-channel visual evoked potential (VEP) measurements of two testees. Theoretical simulations indicate that source orientation and location have major impact on the specificity and therefore they should be taken into account when the optimal bipolar electrode configuration is selected. The results also imply that the new ROISR method bears a strong correlation to the SNR of measurement and can thus be applied in the future studies to efficiently evaluate and optimize EEG measurement setups.

Improved detection of deep sources with weighted multielectrode EEG leads.

The purpose of the present study is to conduct preliminary experimental measurements to validate the improvement in the detection of deep EEG sources achieved with new multielectrode EEG leads. As a measurement we had brainstem auditory evoked potentials (BAEPs), which include deep generators in the brainstem and midbrain. The BAEPs were measured with a 124-channel EEG cap. We have previously developed a multielectrode lead technique, which has its basis in optimization of the sensitivity distribution of a multielectrode lead for detecting signals generated by deep sources. The purpose of the present study is to validate with experimental measurements the results previously obtained with theoretical approach and simulations. The results show that the amplitude SNR of BAEPs obtained with multielectrode lead is on average 1.6 times that of traditional bipolar BAEP lead. Though improvement obtained in experimental measurements is smaller than was theoretically approximated it encourages for further development of the multielectrode leads.