Kernel Current Source Density Method

The last few years have brought dramatic developments in technologies and availability of multi-channel electrodes allowing simultaneous recording of electric potential from many places in the brain. Hence the growing needs for new methods for analyzing ofsuch data.

The choice of a method depends on the signal frequencies one wants to analyze. High frequencies of the extracellular potential reflect electric activity of individual neurons. Our method is mostly dedicated to analyzing low frequencies usually called the Local Field Potential (LFP). LFP is thought to reflect synchronized information processing in dendrites.

Extracellular potential is generated by transmembrane currents – the flow of ions moving inside and out of the neural cells. The long-range nature of the electromagnetic field leads to the effects of local current activity being visible at multiple electrodes what complicates the interpretation of LFP. This is why it is often convenient to reconstruct the current source density (CSD), which defines the profile of transmembrane currents. CSD is positive, where positively charged ions on the average flow into neurons and negative in places where they flow out.

The earlier methods of CSD reconstructions shared two significant drawbacks: they did not allow for reconstruction of sources from recordings on irregularly placed electrodes and were vulnerable to noise in data. Our method, thanks to the application of advanced statistical methods outperforms the preceding methods in the presence of noise. Also, it does not require any regularity in the electrode positions. This gives the experimenter a variety of new possibilities.

Date of publication
17 July 2012