No signal could be collected from a healthy culture confirmed by MCS system

[Edwin Yi]

Hi, neurorighters:

We still can't get signal from our system :(. This time we confirmed the culture is healthy enough to generate signals in a commercial MCS system. The picture is shown below. After that we used NR system to repeat the testing, but no signal could be observed (the screen shot is shown below).

Besides no signal could be oberved, we summerized following questions.

1. For the NR software, we set the parameters according to the previous paper (frontiers in Neural Circuits), but they were not explained in details in the paper. Does anyone have the detailed information, especially for the spike detection and filters?

2. From the figures, many spikes have crossed the threshold but no waveforms could be generated. And since the waveforms represent the effective spikes, is this caused by our settings? (Threshold:5; Pre-spike time: 0.5ms; Post-spike time: 1.5ms; Noise estmation algorithm: adaptive; Projection type: PCA; Projection dimension: 2; Max. K: 4; P-value: 0.01.) When we do the spike hoard, no spike was detected.

3. From the MCS system, the signal is about +/- 20uV. Could I know your amplitude of your typical signals?

4. Since the time window in NR is only 250ms, it may be difficult to obersve the bursts in the display. Could we change the time window?

Thank you!

Edwin

[Jon Erickson]

Hi Edwin,

 

We still can't get signal from our system :(. This time we confirmed the culture is healthy enough to generate signals in a commercial MCS system. The picture is shown below. After that we used NR system to repeat the testing, but no signal could be observed (the screen shot is shown below).

​Sorry to hear you are still having trouble.  Take heart- the system can work very well...at least that's been my experience.  Grounding is really, really important to get a clean signal.  Maybe something to look into.​ Also, have you trialed with SALPA on and off.  SALPA can be great, but one somewhat undesirable property is that it zeros out the signal when it can't make a good fit---this can lead to confounding interpretation of signals sometimes.  For example, if you have a big power spikes (e.g. from incubator) then SALPA may zero out the trace for a while before it recovers.

 

1. For the NR software, we set the parameters according to the previous paper (frontiers in Neural Circuits), but they were not explained in details in the paper. Does anyone have the detailed information, especially for the spike detection and filters?

​This is usually heuristic, detection threshold usually set to something like 5 or 7 x noise typically works reasonably well.  Filter settings are based on the width of the spike waveform.​ Sideband cutoff frequencies should be set to reasonable values to get rid of low frequency drift, keep the ~1ms waveform, get get rid of high freq noise.  So something like a few Hz to a  1 kHz should be sensible.

 3. From the MCS system, the signal is about +/- 20uV. Could I know your amplitude of your typical signals?

​More typical is about 50-100 uV, but some are as small as ~ 20 uV.​

​Hope this helps. Good luck,

Jon​

--
You received this message because you are subscribed to the Google Groups "NeuroRighter Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neurorighter-us...@googlegroups.com.
To post to this group, send email to neurorigh...@googlegroups.com.
Visit this group at https://groups.google.com/group/neurorighter-users.
For more options, visit https://groups.google.com/d/optout.

[Riley Zeller-Townson]

Eric- from what I can see on your traces, I am guessing that the issue is primarily electrical right now.  Your noise appears low, but so low (and given the temporal characteristics) that I would guess there is some sort of connectivity issue with the wiring from the NI daqs to the preamplifier.  The temporal characteristics you show on 57 and 64 are closer to what you would expect to see on all the channels, but with higher rms values for recording electrodes.  

You can see example traces here: https://sites.google.com/site/neurorighter/user-manual/recording-neural-signals

To add to Jon's point on SALPA: neurorighter also uses some waveform shape requirements, in addition to the threshold-crossing criteria, to validate spikes.  As such, not every threshold crossing will be counted as a spike- but threshold crossings that don't survive the validation process are probably due to noise.  You can read about them in more depth in the supplemental section of the fronteirs paper- http://journal.frontiersin.org/file/downloadfile/26827/octet-stream/Supplementary%20Material.PDF/11/5/32539

The settings you describe should not impact detection, beyond the 5xRMS threshold (which is a fine value to use).  

I would try saving raw traces and examining the recording in matlab to see if the raw signal gives a better indication of the noise you have at your electrodes, and then go from there.  Good luck!

--

Riley Zeller-Townson

PhD student, Butera lab

Laboratory for Neuroengineering

Georgia Tech/Emory University

Atlanta, GA, USA