Issue 66 in neurorighter: Spike detection thresholds near zero if use SALPA with stimulation at the beginning of recording protocol.

[neuror...@googlecode.com]

Status: New
Owner: ----
Labels: Type-Defect Priority-Medium

New issue 66 by Silvia.D...@gmail.com: Spike detection thresholds near zero
if use SALPA with stimulation at the beginning of recording protocol.
http://code.google.com/p/neurorighter/issues/detail?id=66

What steps will reproduce the problem?
1. Load stimulation protocol that has stimulation within first 10 sec
2. Train SALPA
3. Run protocol

What is the expected output? What do you see instead?
Noise levels were much lower than expected (near zero).

What version of the product are you using? On what operating system?
v1.0.2.521 on Windows XP

Please provide any additional information below.

[Jon Newman]

Hi Silvia,

This is actually not a bug. The way spike detection works is that NeuroRighter looks at the first 10 seconds of data from each channel and calculates the V_RMS for each using 100 ms windows. It then makes  a histogram of the RMS measurements from each window, for each channel. The detection RMS value is value corresponding to the lowest 10% of RMS measures. This ensures that bursts or spikes that occur during the estimation period do not make the RMS value artificially high.

Because SALPA is blanking during this period, it is making a bunch of the windows have an RMS of 0. Thus, the lowest 10% of RMS values correspond to 0, and the threshold is set to 0. 

A work around is to simply start your stimulation once the training period is over (10 seconds in).

-Jon

--
You received this message because you are subscribed to the Google Groups "NeuroRighter Users" group.
To post to this group, send email to neurorighter-users@googlegroups.com.
To unsubscribe from this group, send email to neurorighter-users+unsub...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/neurorighter-users?hl=en.

--
Jon Newman
Wallace H. Coulter Department of BME
Georgia Tech and Emory University
Cell: (678) 431-1697
URL: www.prism.gatech.edu/~jnewman6

[Riley Zeller-Townson]

Currently this is the case even if the spike detector is 'trained' prior to starting the stimulation protocol.  I think in that case (where the user has explicitly trained the detector prior to starting the protocol) it might be more intuitive to use the noise levels from the training period, rather than the first 10 seconds of the current run.  Thoughts?

To post to this group, send email to neurorigh...@googlegroups.com.
To unsubscribe from this group, send email to neurorighter-us...@googlegroups.com.

For more options, visit this group at http://groups.google.com/group/neurorighter-users?hl=en.

--

Riley Zeller-Townson

PhD student, Potter lab

Laboratory for Neuroengineering

Georgia Tech/Emory University

Atlanta, GA, USA