Neural coding deals with the problem on how neurons or a network of neurons processes a stimulus and creates a response in form of electrical action potentials. Information then might be encoded in the rate of the action potentials or their timing.

My questions now are:

  1. Do nerves, e.g. in the arm of a human, merely "relay" any information, or do they also process incoming signals in any way? As a result, the electrical signal would then differ when measured at the beginning and the end of a nerve.

  2. Can a nerve be stimulated at one point with a certain electrical signal as input (with varying rate etc. to encode information) and can the output be measured at another point?

If I understood correctly, this is done at a electroneuronography (ENoG) only that the electrical stimulus is not varying in rate or other properties.

Any links to research that deals with similar things would be great.

  • $\begingroup$ Hi Eser, welcome at CogSci. Would a reflex count as processing a stimulus? When you stimulate the skin with heat or pressure (or pain), for instance, it could generate a reflexive moment which is not controlled/processed by the brain. $\endgroup$ Jun 23, 2017 at 9:46
  • $\begingroup$ Thanks for your comment Robert. Yes, I think a reflex would prove that the nerves are somehow processing the stimulus. What I am particular interessting in, and which may sound a bit strange, is if one could excite a nerve with an electrical stimulus that encodes information at one point and pick up the response of the nerves at a different location. It doesn't matter HOW exactly the nerve structures process the information but the response of cause needs to depend on the stimulus. Like in a black box system with an unknown transfer function. $\endgroup$
    – Eser
    Jun 25, 2017 at 12:24

2 Answers 2


Short answer

  • Peripheral nerves can be stimulated by, e.g., electrical pulses and their responses can be recorded back by using recording electrodes.
  • Peripheral nerves generally do not process the signal. However, they can, and likely will, behave in a nonlinear manner, meaning that their transfer characteristics may change over time, depending on the stimulus.

I'll approach this question based on the cochlear implant (CI). In short, these devices are implanted into the inner ear in people with pronounced hearing loss (Fig. 1). CIs stimulate the auditory nerve directly with electrical pulses, thereby bypassing the degenerate hair cells and restoring functional hearing.

The auditory nerve responses elicited by CIs are routinely recorded to assess implant functionality. These recordings are called electrically evoked compound action potential (eCAP) responses (Fig. 2). Basically, one electrode is stimulated and the nerve response is measured by another electrode. All the action potentials of the individual auditory nerve fibers are basically recorded synchronously, resulting in a compound action potential (CAP). This will probably answer your second question, namely yes, nerve responses can be elicited electrically and recorded back with the proper recording setup.

To go back to your first question - nerve responses are nonlinear, as neural responses show adaptation to the stimulation. Basically, nerves get tired of firing after a while. To go back to the CI example; if high rates of electrical stimuli are applied to the inner ear, the auditory nerve shows a slow adaptive behavior, causing the nerve response amplitude to decline slowly over time, this is caused because more and more fibers become less responsive. Then there is a short-term effect caused by refractoriness of individual fibers. A few milliseconds after a fiber has fired, it is unresponsive to a following stimulus. This causes some fibers to skip the stimulus. During a high-rate pulse train, this results in e eCAP to alternatingly have a normal and low amplitude (Fig. 3). This answers your first question, in a nutshell, namely that nerves can alter the input signal for sure. The process of adaptatin is widespread in the periphery, for example all sensory systems show it to some degree. It is, however, not a real processing such as the processing happening in the peripheral retina for example where complex circuitry markedly alters retinal output to the optic nerve.

Fig. 1. Cochlear implant. source: Speech Buddy

Fig. 2. Princple of eCAP recording. source: A.D. Costlow

pulse train
Fig. 3. eCAP amplitudes during high-rate (900 pulses/s) stimulation of a CI. figure excerpt adapted from: Hughes et al. (2014)

- Hughes et al., Hear Res (2014); 316: 44-56

  • 1
    $\begingroup$ Thanks for the explanation AliceD. One quick follow up question though: Could the stimulation and recording on some nerves also be achieved with a non-invasive setup? If yes, what kind of equipment would you need for that? $\endgroup$
    – Eser
    Jun 28, 2017 at 10:24
  • $\begingroup$ @Eser I think that's difficult. Here is a link to a minmimally invasive procedure to record mechanoreceptors in the skin $\endgroup$
    – AliceD
    Jun 28, 2017 at 18:20
  • $\begingroup$ Okay but what is then done in electroneuronography (ENoG), isn't that what is happening there? Or are the measurements of ENog so coarse that you can only measure the time difference between stimulation and response but not the exact form of the action potentials? $\endgroup$
    – Eser
    Jun 29, 2017 at 9:30
  • $\begingroup$ @Eser - I am not too familiar with ENoG, but by the looks of it CAPs can be used to record it. Also muscle activity (EMG) can be recorded back. $\endgroup$
    – AliceD
    Jun 29, 2017 at 10:24

Does the peripheral nervous system processes information like central neurons do?

No. Central neurons predominately process information through neural networks. A neural network is basically a group of neurons who have synapses on each other thereby providing excitatory and inhibitory inputs. The peripheral nervous system is generally thought of as being a relay between the periphery (e.g., limbs) and the brain. They do not generally form synapses with neurons outside the CNS and therefore cannot perform processing like neurons in the CNS.

Do nerves, e.g. in the arm of a human, merely "relay" any information, or do they also process incoming signals in any way? As a result, the electrical signal would then differ when measured at the beginning and the end of a nerve.

At the individual neuron level the "signal" is generally thought of as being an action potential. Action potentials are typically described as all-or-none events and propagate along the nerve fiber. The exact morphology of the action potential can change as it propagates, but when it gets to the end, it will cause the release of neurotransmitter to signal the next neuron.

An alternative definition of "signal" are post synaptic potentials. These are the small changes in voltage that a nerve fiber undergoes do to the presence of neurotransmitters. The Hodgkin and Huxley model provides a mathematical description of how this process works.

Can a nerve be stimulated at one point with a certain electrical signal as input (with varying rate etc. to encode information) and can the output be measured at another point?

No (well maybe). Stimulating a nerve at a particular point is difficult. Hodgkin and Huxley in their Nobel prize winning work injected current (maybe voltage) into the giant axon of the squid and measured the response. The technique of intracellular recordings involves patch clamping and is not easy (again a Nobel prize worthy discovery). Patch clamping a cell in two places is probably theoretically possible, but I pity the grad student who has to do it.

With extracellular recordings it is difficult to know "where" you are stimulating and recording. It is often hard to even know what neuron you are recording from.

  • $\begingroup$ OP is not talking about single-fiber action potential recordings, but gross compound action potential recordings. $\endgroup$
    – AliceD
    Jun 28, 2017 at 18:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.