Triggering Workflow and Noise Reduction

[Paolo Spaiardi]

Dear Open Ephys Team,

We purchased an Open Ephys Starter Bundle equipped with an RHD 32-Channel Stim/Recording Headstage (RHD2132), to be interfaced with 4-contact cuff electrodes (Microprobes; model NC-0.6-4-100P-1-0.5-SUT-300-CONN).

I am writing to seek your technical advice on how to best configure our experimental protocol.

Our goal is to record bioelectrical activity (NCAP) from the rat facial nerve in response to an odor stimulus delivered by an Air Puff Whisker & Olfactory Stimulator (LabeoTech). Simultaneously, we will record whisker movements using an infrared Behavioral Camera (LabeoTech).

To optimize data acquisition, we would like to synchronize the electrophysiological recording, the video acquisition, and the puff delivery. Specifically, we would like to configure the Open Ephys GUI so that starting the recording automatically triggers:

1.    An instantaneous TTL pulse to activate video acquisition.

2.    A second TTL pulse, after a specific delay (e.g., 5 seconds), to trigger the odor puff delivery.

Regarding this setup, I have the following questions:

·       Is it possible to set up this type of workflow directly within the GUI?

·       Which specific plugins would you recommend using (e.g., Event Broadcaster, Pulse Pal, or others). Is it possible to use just Network Events and avoid the use of Arduino ?

·       Can we utilize the two I/O boards included in our bundle, or is it necessary to use an external microcontroller (e.g., Arduino)? If so, could you provide guidance on the recommended configuration?

During preliminary recordings, we noticed significant noise in the signal. Our current configuration is as follows:

• 3 headstage channels are connected to 3 of the 4 cuff-electrode contacts.
• The 4th contact is connected to the headstage REF.
• The remaining 28 unused channels are bundled together and shielded with a metallic mesh connected to the Faraday cage ground.

Does this configuration seem correct to you, or would you suggest an alternative referencing strategy or a different way to handle unused channels to improve the signal-to-noise ratio?

Thank you in advance for your time and support.

Best regards,

Paolo

[Josh Siegle]

Hi Paolo,

I just responded on Discord, copying the message here as well:

If you can tolerate 10s of milliseconds of jitter in your output pulse timing, this setup is possible using just the Network Events plugin and the Acq Board Output plugin. You won't need any additional devices. First, make sure one of your I/O boards is connected to the digital output port (2nd HDMI port from right) and one is connected to the digital input port (1st HDMI port from the right) Next, build a signal chain that (at least) includes the following plugins: Acquisition Board > Network Events > Record Node > LFP Viewer > Acq Board Output # 1 > Acq Board Output #2 Next, write a Python or Matlab script that:

1. Uses an HTTP server command to start recording
   
2. (after a short delay) Sends a Network Events message with a TTL event that triggers Acq Board Output #1 (mapped to the digital output line connected to your camera system). Ideally the camera frame TTL events will be acquired via the Acquisition Board's digital input port so you know exactly when each frame was acquired.
   
3. (after 5 seconds) Send a Network Events message with a TTL event that triggers Acq Board Output #2 (mapped to the digital output line connected to your odor puff system). This digital output should be split so it can also be acquired by one of the Acquisition Board's digital inputs.
   
4. (after however long you want to record) Uses an HTTP server command to stop recording
   

The only downside to this is that triggering events this way will lack sub-millisecond precision, because everything is processed in software. If you need very precise repeatability across sessions then you'll need a hardware-based event triggering system like an Arduino or a Pulse Pal -- but it sounds like that's not needed here.

Regarding your question about noise, I don't have any experience with cuff electrodes, so I'm not sure about the ideal referencing strategy. Have you tried adding a Common Avg Reference plugin in the GUI, prior to visualizing the data in the LFP Viewer? Often that's sufficient to clean up noisy signals.

Best,

Josh

--
You received this message because you are subscribed to the Google Groups "Open Ephys" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-ephys+unsubscribe@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/open-ephys/af87bee9-f253-487d-82f5-4ab7cea16473n%40googlegroups.com.