====== Extracellular Recording of Neural Activity ======

Contributed by [[flavio@salk.edu|Flavio Frohlich]]

Action potentials result in a transient change in the extracellular space which can be recorded with microelectrodes. The key advantages of extracellular recording are:
  - Relatively simple set-up (in comparison to intracellular recordings)
  - Under certain conditions, one can record from a population of neurons with single cell resolution.

The key disadvantages are:
  - There is no perfect way to separate all recorded spikes into different cells (spike sorting).
  - Very little can be said about the type of neuron since the intracellular firing patterns in response to stimulation cannot be determined. Also there is no way to fill and therefore label the cells recorded from.

Here, we will walk through the following steps of an extracellular single unit recording:
  - Choice of equipment (electrodes, amplifiers)
  - Data processing (spike sorting).

Later on, more details and technical issues will be covered.


===== Choice of Equipment =====

==== Electrodes ====
Most extracellular recordings are performed with metal microelectrodes. Recently, silicon probes have become popular, also.
=== Metal Microelectrodes ===
Metal electrodes are basically a metal wire which is insulated except for the tip which will have some type of geometry. Many parameters influence the behavior of the electrodes, unfortunately there is no definite mapping from parameters to recording properties nor is there a solid theory of how extracellular spikes are picked up my metal microelectrodes. Thus, key goal of this document is to collect the experimentalist's experience to build some kind of collective understanding of what electrode to pick for a given purpose. Also, individual manufacturers will be discussed because some of them have very absurd business models which you want to be aware of before you plan on using their electrodes. This is a friendly way of me to say that some have astronomical prices, abysmal quality, and waiting lists order of magnitude many months, at least.






===== Source of Extracellular Signal =====
While extracellular recordings are an established techniques in many in vivo labs, there is quite some uncertainty about the origin of the signal recorded by extracellular electrodes. This not just a technical detail since we still lack a good understanding of the possible biases of extracellular recordings. Only understanding the source of signal will eliminate doubts and answer questions related to the bias issue.

Essential, no charge is transfered between the extracellular space and the metal electrode. Rather, the recording is a "capacitative" recording, essentially indicating that the extracellular field generated by ion flux moves the electrodes in the metal electrode and thus creates the signal that is then amplified and recorded.


===== Spike Sorting =====
Spike sorting refers to the procedure that assigns cell identity to action potentials that were recorded with extracellular electrodes. Bascially, the procedure can be described as a three step process:
  - Collecting presumed action potentials,
  - Clustering all presumed action potentials into sets that correspond to different cells.
  - Assessment/quantification of spike sorting quality.

In other words, you start with a heap of negative-positive deflections that correspond to action potentials from active neurons close to your recording electrodes (multiunit data) and you want to end up with knowing which spikes came from which cell ("single-unit" resolution).

It is curcial to understand that there is NO single best methode for spike sorting and that all approaches I know of include a manual / experience-based component that has so far eluded a formal mathematical description. Unfortunately, it is therefore important to understand at least the key concepts to be able to properly intprete results derived from extracellula recordings.