[Eeglablist] Problems with a CHEPS thermode and EEG reading

[Malte Anders]

Dear eeglab-list,

I have solved the problem on my own.

For some reason, everything works as it should be once you connect the
grounding connector of the PATHWAY thermostimulator to the subject (yes, to
the subject! We will be using an antistatic wrist strap for a few bucks in
the future that is connected to the thermostimulator's grounding port). It
didn't work when I wired that connector to a metallic radiator or any other
object but to the subject itself.

Sounds weird, but works. Just wanted to let everybody know! And thanks to
everybody who tried to help me!

Malte

2017-12-13 11:04 GMT+01:00 Malte Anders <malteanders at gmail.com>:

> Dear EEG lab list,
>
> I will try for a third and last time to get some help here (thanks to
> anybody who tried to help so far!). I have now identified the culprit to be
> the CHEPS/thermostimulator. This happens:
>
>
> Subject is undergoing EEG reading, everything is fine (range of the EEG is
> approx. up to 100 microvolts as it should be). I have eliminated nearly all
> electromagnetical interference. I even connected the thermostimulator to an
> outlet in another room.
>
> As soon as the subject touches the kapton foil of the CHEPS thermode, the
> eeg shows a sinus-like rhythm which should be approx. 50 Hz. The signal
> then goes up to approx. 400 microvolts.
>
> *The sinus rhythm is even visible when I apply a notch filter between
> 48-52 Hz AND a bandpass from < 1 Hz and > 30 Hz.*When there is a heat
> stimuli up to 51 °C, the signal does what is shown in this image (
> https://imgur.com/a/5pcBT) before and after the red trigger line and the
> peak goes up to 1-2 mV,
>
> *even when the signal is filtered.*
> The only way to remove this pattern is re-referencing (obviously) but I am
> getting very mixed results, ICA is impossible to perform and thus I am
> afraid that this is messing up my whole evoked potential EEG reading.
>
> When there is 5 mm (or more) distance between the electrode and the
> subject, the EEG is totally fine and in the 100 microvolt range, but
> oviously the subject needs to be in contact with the thermode for my setup
> to work...
>
>
>
> I found this regarding CHEPS noise: https://www.ncbi.nlm.nih.gov/
> pmc/articles/PMC3881311/
> This sentence points out the problem pretty well: The CHEPS artifact is
> highly dynamic, yet deterministic, switching rapidly from one frequency to
> another, with different spatial components. The galvanic connection between
> the subject and the CHEPS probe alters its performance, making
> pre-characterization difficult.
>
>
> Is there anybody who has done evoked potentials/CHEPS EEG readings before
> and can help me with this? Or can I simply ignore the artifact,
> re-reference and continue like usual? I had readings where I definitely saw
> evoked potentials but others which where so messed up I didn't see a
> thing...
>
> Thank you!
> Malte
>
>
> 2017-11-23 8:44 GMT+01:00 Malte Anders <malteanders at gmail.com>:
>
>> Dear EEG Lab list,
>>
>> I dug a bit deeper and my nightmares have come true... at least now I
>> know what is going wrong. See this screenshot:
>> https://imgur.com/a/5pcBT
>>
>> The EEG is unreferenced, thats why its in the 1 mV range. The blue line
>> marks the spot where the subject was handed the CHEPS electrode. Before
>> that, somebody else was holding the electrode. As you can see, the signal
>> was perfectly clean before the subject took the thermode in her hand. She
>> _only_ touched the outer metal case of the thermode. Filters can be seen in
>> the top row (bandpass < 1Hz, > 60Hz, notch 48-52 Hz). We're in Europe so
>> notch is between 48-52, those filters were just applied for visualization
>> purposes.
>>
>> After the thick blue line, there is some random noise in the line which
>> can't be 50 Hz as i have notch filtered the EEG! I am currently out of
>> ideas but still pretty happy as I found the culprit for my messed up
>> data... The signal is gone after re-referencing e.g. to earlobes - which
>> makes sense - but a noise like that should not be in an EEG whatsoever so I
>> need to solve this problem.
>>
>> The red line (Trigger 9) is a 51 °C CHEPS stimulus and this is the
>> pattern I get every time I am applying a stimulus. Before the trigger: up
>> down up down and then after the trigger: 0,5 seconds of basically nothing
>> (flat line which, if zoomed in, shows a sinus wave as seen in the first
>> post). As this is the crucial part for me (obviously) this can't be a good
>> sign...
>>
>> The thermode looks like this:
>> https://imgur.com/a/nAOAk
>>
>> The CHEPS machine looks like this:
>> https://imgur.com/a/0OWQi
>>
>> The red circle marks the _only_ connection to the EEG which is a 5V
>> trigger signal. Even if this is disconnected I get a really weird EEG
>> reading as seen in the first screenshot so this can't be a problem. There
>> are _no_ other physical connections besides the subject who is holding the
>> thermode.
>>
>> The EEG ist +5 V and draws 6A Maximum. Both are connected to a power
>> outlet (230 Volts, 50 Hz) in the same room. I still have no idea if this
>> makes a difference though.
>>
>> I am happy for any input here. Thank you so much!
>> Malte
>>
>>
>> 2017-11-20 11:30 GMT+01:00 Malte Anders <malteanders at gmail.com>:
>>
>>> Dear EEG Lab list,
>>>
>>> we are using a CHEPS thermode on a Medoc Main Station (
>>> https://medoc-web.com/products/pathway-model-cheps/) and a 64 Channel
>>> g.HIamp from gtec with active electrodes (basically electrodes with an
>>> amplifier in it). I am currently measuring heat evoked potentials, so
>>> nothing really fancy.
>>>
>>> When doing the eeg recordings, I am experiencing a strange behaviour I
>>> can't make heads and tails of. I am usually looking at an unreferenced EEG
>>> signal while recoding as we are doing and average referecing in the
>>> post-processing. Here is a screenshot of how the eeg looks _every_ time a
>>> trigger signal (CHEPS) of 51°C is applied to a subject (non-filtered, non
>>> referenced)
>>> https://imgur.com/ANxIPQb
>>>
>>> Filtering the data doesn't make it better (notch, bandpass), just
>>> smoother, re-referencing (either average or earlobe, doesn't matter)
>>> completly removes the very significant pattern and results in a what I
>>> would call "normal" EEG.
>>>
>>> I am especially worried about the very "flat sinus wave" after the
>>> trigger (trigger = peak temperature of 51 °C). When I measure an eeg
>>> without a subject and start the CHEPS electrode, no such pattern is seen.
>>> What I haven't tried is measuring and EEG on a subject, start the CHEPS
>>> measurement without the electrode applied to the subject to see if this
>>> pattern reoccurs, but I will try that. The only physical connection between
>>> EEG and CHEPS station is a +5V trigger cable which hopefully doesn't impose
>>> any electrical disturbances.
>>>
>>> I just want to know if I need to be worried about the flat sinus wave
>>> after the trigger. I know that it is gone after the re-referencing, but I
>>> am not sure if there is any electrical interference I am currently not
>>> paying attention to. Also the pattern before the trigger always looks
>>> exactly the same (up, down, up, down and then flat sinus wave)!!! As I am
>>> looking for potentials in a 5-10 microvolt range I just want to be extra
>>> cautious, and the unreferenced signal produces this pattern in the
>>> millivolt-range, that's why I am a bit worried.
>>>
>>> I am hoping this is the "normal" physical response of a subject in a
>>> heat evoked potential, but I just wanted to double check with somebody who
>>> has done evoked potentials before to rule out any interference here.
>>>
>>> Thank you so much!
>>> Malte
>>>
>>>
>>>
>>
>>
>> --
>> Mit freundlichen Grüßen,
>>
>> Malte Anders
>>
>
>
>
> --
> Mit freundlichen Grüßen,
>
> Malte Anders
>



-- 
Mit freundlichen Grüßen,

Malte Anders
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20171220/cbb023d0/attachment.html>