Dear Kristie,

I agree with the other persons who have answered your question. Basically the electrode in the brain records the signal in the extracellular medium which is a superposition signal of LFP and unit activity. The separation in LFP (local field potentials) and action potentials or single unit activity (SUA) is normally done in the amplifier system by different filters as reported by William. Normally the LFP signal has a larger amplitude as the action potentials of the single or multi unit activity. Therefore the LFP requires a low pass filter as well as a smaller amplification in comparison to unit activity which has a much smaller amplitude and requires a bandpass filter (e.g. 300Hz…6kHz) to remove the low frequency components. I would like to point out that the electrode is important for the recording quality. As you probably know metal microelectrodes have a strong capacitive component of the complex electrode impedance. A recording electrode that is optimized for single unit isolation normally has a small exposed metal area and a high electrode impedance in the order of some MOhm (e.g. 2-3Mohm, normally measured with sine wave test signal at 1kHz). Due to the capacitive nature of this kind of electrode the impedance value is much higher for lower frequencies (as the LFP). Therefore it is highly recommended to use a good shielded high impedance input preamplifier system which records the complete signal and the frequency separation is done in the main amplifier filter system. Or you use a low impedance electrode that picks up multi unit activity. This electrode has a lower impedance in the LFP range compared to a single unit electrode and therefore is not so sensitive for 50Hz or 60Hz interference. In case of multi-unit recording I would recommend to use tetrodes. Let me know if you have any further question. Best wishes, Dirk

Yes, this can be done at the same time. Classically, single unit recordings at performed with a bandpass filter with a pass band of something like 100 - 2000 Hz, while LFP recordings are performed with either a low pass filter at < 100 Hz, or a band pass with a very low cut off frequency (0.1Hz) and an upper limit of < 100Hz.

So what you do it record with either a low-pass filter set at around 1000/2000 Hz, or a bandpass filter from something very low, up to 1000/2000 Hz. Then you can digitally filter the signal afterward, separating the two frequency components.

You have to separe the amplitude modulated wave (LFP) from the spike train (axonal firing of a single neuron).

William is right: spike and LFP activity can be recorded on the same electrode. Moreover, If you want to look at spike - LFP  interactions, for example spike-LFP timing, I'd recommend you record your data with your analog filters "wide open", 0.1~5kHz (sample at at least 10kHz) and later separate the two signals using digital bandpass filtering (offline).

These all is right provided that your single electrode is a tetrode consisting of 4 wires. Only in this case you will be able to isolate spikes from individual neurons ("single units"). If you record LFP and spiking activity with a single wire, you can analyze only, so called, multi-unit (MUA) spiking, i.e. a mixture of spikes from many neurons.

Dear Kristie,

I agree with the other persons who have answered your question. Basically the electrode in the brain records the signal in the extracellular medium which is a superposition signal of LFP and unit activity. The separation in LFP (local field potentials) and action potentials or single unit activity (SUA) is normally done in the amplifier system by different filters as reported by William. Normally the LFP signal has a larger amplitude as the action potentials of the single or multi unit activity. Therefore the LFP requires a low pass filter as well as a smaller amplification in comparison to unit activity which has a much smaller amplitude and requires a bandpass filter (e.g. 300Hz…6kHz) to remove the low frequency components. I would like to point out that the electrode is important for the recording quality. As you probably know metal microelectrodes have a strong capacitive component of the complex electrode impedance. A recording electrode that is optimized for single unit isolation normally has a small exposed metal area and a high electrode impedance in the order of some MOhm (e.g. 2-3Mohm, normally measured with sine wave test signal at 1kHz). Due to the capacitive nature of this kind of electrode the impedance value is much higher for lower frequencies (as the LFP). Therefore it is highly recommended to use a good shielded high impedance input preamplifier system which records the complete signal and the frequency separation is done in the main amplifier filter system. Or you use a low impedance electrode that picks up multi unit activity. This electrode has a lower impedance in the LFP range compared to a single unit electrode and therefore is not so sensitive for 50Hz or 60Hz interference. In case of multi-unit recording I would recommend to use tetrodes. Let me know if you have any further question. Best wishes, Dirk

Metal electrodes are not good at recording slow signals.

Two suggestions:

(1) a 'reversible' metal electrode (treated Platinum tip)

(2) A glass pipette of a few Mohm impedance.

There is an old literature on construction of these kinds of electrodes.

I absolutely agree with William Nickell that metal electrodes are not optimal for recording local field potentials. But the question was if one can record single unit activity AND local field potentials using one electrode. Basically I would say YES but it is a compromise. If one uses a platinum plated low impedance metal electrode it will be possible to get local field potentials and simultaneously multi-unit-activity. If one wants to record local field potentials and simultaneously single-unit-activity it is recommended to use a tetrode as I suggested in my previous answer. The tetrode is 4 contact, low impedance electrode. Usually the 4 metal contacts of the tetrode are plated to lower the impedance of each of the four metal contacts. This electrode picks up multi-unit- activity. To isolate single units from this multi-unit recording one has to use a tetrode spike sorter which is based on the stereotrode effect (for details see: McNaughton BL, O'Keefe J, Barnes CA. The stereotrode: a new technique for simultaneous isolation of several single units in the central nervous system from multiple unit records. J Neurosci Methods. 1983 Aug;8(4):391-7). All of the four tetrode conductors also will record field potentials. Each tetrode contact records the superposition signal from the extracellular space (unit activity & field potentials). After pre-amplification the superposition signal is separated by filters in LFP and Unit activity and amplified again and finally digitized with a data acquisition system. Finally the multi-unit activity is analyzed with a tetrode spike sorter to isolate the single units from the multi-unit recordings (for details concerning tetrode spike sorting see: Gray, C.M., Maldonado, P.E., Wilson, M. & McNaughton, B.; Tetrodes markedly improve the reliability and yield of multiple single-unit isolation from multi-unit recordings in cat striate cortex. J. Neurosci. Methods 63, 43−54).

You will absolutely need to band pass filter your data.  You can separate spikes recorded with single electrodes into clusters based on the principal components of the waveforms (no need to use tetrodes if you use an automated Expectation Maximization algorithm such as KlustaKwik).

http://klustakwik.sourceforge.net/