Evoked Potentials - Science topic

Hi Younsoo. I was able to kind of resolve my issue although I had to put a pause on these experiments. The biggest issue for me was the stimulating electrodes used. When I bought fresh ones, I was able to elicit responses. I hope you are able to figure it out!

You can look for papers on this topic and check if the authors share their datasets in public repositories.

Hi, for generate EEG signals, you can directly create a sine wave with the specific frequency and add some random noise.

There is also a link of Mike`s course about "Simulating data to understand analysis methods", hope it can help you!

Using only one electrode can reduce the amount of information you can get, and reduce also the signal processing techniques (like ICA) you use to get a good features.

Good luck.

Why don`t you try non-parametric Mann-Whitney test? In addition, it appears to me you are comparing more than two.groups, and in that case you cannot use t test nor MW. For that matter, non normal.data can be analyzed by kruskal-wallis and a post-hoc (i use nemenyi).

Umema Zafar , If you can be a little more specific, I'll be happy to help!

I totallly agree, technical issues must be ruled out like active electrode and reference changing. This means active is put on the reference site and reference is put on active site. Nevertheless there are programs that allows to change the polarity of the recorded signal, so maybe you can still use that information. But we are open mindd so, rule out and see.

Hi Tenea Welsh ,

If I understood well and in my opinion, for the purpose of statistical analysis and publication, you should perform separate analysis for those neurons treated with picrotoxin (as the GABA antagonist) and those treated with DNQX (as the AMPA antagonist). Although APs seem somehow identical and seemingly inconsequential in some experiments, the mechanism of action and their role in generation and propagation of APs and then fEPSPs or IPSPs are rather different in terms of channel activity and resultant synaptic activity. In particular when you use two antagonists which act differently and on different receptors.

You may analyze the recording and stimulation response of both separately and then compare them against each other though the later depends mainly on your setup, hypothesis/goal and recording/stimulation parameters. This can provide a better and conclusive insight regarding your common goal for different types of receptors and treatments too.

Hey Danesh,

I have not explored this very far, but Jim Cavanagh (and others) has developped this site: predictsite.com. Not sure what is included yet, or whether they will be large enough datasets, but might help.

Ali makes many excellent points on this issue. It is worth noting that he is drawing on the proper definition of transition state in the EEG literature, namely actual critical transition states typically assessed via modeling nonlinear dynamics (e.g. via measures of chaos, e.g. lyapunov exponents) often in epilepsy, seizures and other shifts in conscious mode.

I believe Anoop is referring to the time-interval of an ERP waveform that lies in-between a successive maxima (peak) and minima (trough) as a transition period.

Alas, I am not intimately familiar with the nuances of the N2/P3 complex so I'll abstain from recommending readings in that area. I'm certain you'll be able to find what you need with a quick lit search tgh! Or asking an expert on that topic.

It's absolutely true that the same issues that influence transitions influence peaks and troughs directly too. The problem just gets even more complicated when trying to changes in the "zones" between them.

The observation of an auditory "transition zone" being more stable (on what metrics if you don't mind me asking?) is an interesting one though. Do let me know when you get a paper out on that. I'd love to check it out.

Older but pioneering methodological works.

Whipple, H. E. (1964). Sensory evoked response in man. Annals of the New York Academy of Sciences, 112, 1-546.  [Includes about 20 articles.]

Donchin, E. & Lindsley, D. B. (Eds.) (1969). Average evoled potentials: Methods, results and evaluations. Washington, DC: N.A.S.A. [Incl;udes 7 chapters and 6 supplements.]

you will be welcome !

It would be interesting to quantitate the cortical potentials to wrist flexion torque, wrist extension torque, with prescribe torque objectives, in regard to velocity of torque development, amplitude of torque generated. Then compare rhythmic patterns with  alternation of flexion and extensor movement responding to movement cueing in regard to direction, velocity, and force.

It is very surprising to hear that single pulse TMS on forearm muscle surface EMG is producing a huge artifact. First you need to assure whether it really is a TMS-induced artifact (I doubt that the TMS pulse disturb such a remote recording source on the forearm?).  Second option is that you got a mechanical (stimulator) artifact, as the forespeakers point out it is important to check cable-crossing, grounding, maybe change the socket. Third point might be that the surface EMG is not only recording FDS but more muscle at the same time that is then generating a artifact-like signal, might be likely because your recording on hand muscles are fine. For filtering I refer to the forespeakers, another option I propose is a linear interpolation filter. Good luck!

You can model the sources -with LORETA or any other model- of an ERP (the all potential or a specific window centered in one component) or to an EEG epoch (segment). And for both you can do it in the time domain or in the frequency domain. There is a lot of papers on that. I suggest you to look in them for the methodology.

It is desirable to use constant current and not constant voltage for the stimulation. The impedance (resistance) of the tissue is changing due to some electrochemical processes (polarization, oxydation/reduction). Thus, in constant voltage stimulation, the current delivered to the tissue will be dependent on the resistance of the electrodes and the tissue, simply by the Ohm's law. However, in the case of constant current stimulation, the changes in resistance of the electrode/tissue will have a little effect on the delivered current. Again, according to the Ohm's law, the resistance could increase significantly and diminish the current injected to the tissue. However, one can keep the current constant, and overcome the changes in resistance by adjusting the voltage. Obviously, there is a limit of how much we can increase the voltage to overcome the resistance changes >> stimulation compliance. Most constant current stimulators have a compliance of tens to hundreds of volts.

Pacer Scientific (pacersci.com) has a "Mart" link where people advertise used electrophysiology equipment for sale. Also biosurplus.com runs a lot of used equipment auctions of equipment cast off by industry. I've bought an Axopatch from them and it was fine.

If you are new to the evoked potential methodology, I highly recommend Steve Luck's book on event-related potentials. It does a great job of introducing you to the method, provides lots of practical tips on designing and executing experiments, and useful guidance on how to analyze and interpret your results. Here's a link to the book: