Data Synchronization

Synchronizing behaviour and other experimental events with stimulation or recorded neural data is a fundamental component of neuroscience data collection and analysis. The exercises below will walk you through some common synchronization problems encountered in systems neuroscience experiments, and how to handle them using Bonsai.

Exercise 1: Synchronizing video from two webcams

• Insert a CameraCapture source and set it to index 0.
• Insert another CameraCapture source and set it to index 1.
• Combine both sources using a WithLatestFrom combinator.
• Insert a Concat (Dsp) operator and set its Axis property to 1.
• Insert a VideoWriter sink and record a small segment of video.

How would you test the synchronization between the two video streams?

Note

You can use the FileCapture source to inspect the video frame by frame by setting the Playing property to False. After setting the FileName property to match your recorded video, run the workflow, open the source visualizer, and then right-clicking on top of the video frame to open up the seek bar at the bottom. You can use the arrow keys to move forward and back on individual frames.

Reaction Time

For this and subsequent tutorials, we will use a simple reaction time task as our model systems neuroscience experiment. In this task, the subject needs to press a button as fast as possible following a stimulus, as described in the following diagram:

The task begins with an inter-trial interval (ITI), followed by stimulus presentation (ON). After stimulus onset, advancement to the next state can happen only when the subject presses the button (success) or a timeout elapses (miss). Depending on which event is triggered first, the task advances either to the Reward state, or Fail state. At the end, the task goes back to the beginning of the ITI state for the next trial.

Exercise 2: Generating a fixed-interval stimulus

In this first exercise, you will assemble the basic hardware and software components required to implement the reaction time task. The wiring diagram below illustrates the hardware assembly. You can wire the LED into any digital input pin, but make sure to note the pin number for the steps below.

We will start by using a fixed-interval blinking LED as our stimulus.

• To configure the Arduino analog sampling rate, insert a CreateArduino source.
• Configure the PortName to the Arduino port where the microcontroller is connected.
• Configure the SamplingInterval property to 10 ms.

• Insert a Timer source and set its DueTime property to 1 second.
• Insert a Boolean source and set its Value property to True.
• Insert a DigitalOutput sink and set its Pin property to the Arduino pin where the LED is connected.
• Configure the PortName to the Arduino port where the microcontroller is connected.
• Insert a Delay operator and set its DueTime property to 200 milliseconds.
• Insert a Boolean source and set its Value property to False.
• Insert a DigitalOutput sink configured to the same Pin and PortName.
• Insert a Repeat operator.

Exercise 3: Measuring reaction time

• Insert an AnalogInput source.
• Set the Pin property to the analog pin number where the duplicate LED wire is connected.
• Insert a second AnalogInput source.
• Set the Pin property to the analog pin number where the button is connected.
• Connect both inputs to a Zip operator.
• Insert a CsvWriter sink and configure the FileName property.
• Insert a RollingGraph visualizer and set its Capacity property to 1000.
• Run the workflow, and verify that both the stimulus and the button are correctly recorded.

Exercise 4: Synchronizing video with a visual stimulus

To analyze movement dynamics in the reaction time task, you will need to align individual frame timing to stimulus onset. To do this, you can take advantage of the fact that our simple visual stimulus can be seen in the camera image and recorded together with the behaviour.

• Starting from the workflow in the previous exercise, insert a CameraCapture source and position the camera such that you can see clearly both the LED and the computer keyboard.
• Insert a VideoWriter sink and configure the FileName with a path ending in .avi.
• Insert a Crop transform and set the RegionOfInterest property to a small area around the LED.
• Insert a Grayscale transform.
• Insert a Sum (Dsp) transform. This operator will sum the brightness values of all the pixels in the input image.
• Select the Scalar > Val0 field from the right-click context menu.
• Record the output in a text file using a CsvWriter sink.
• Open both the text file containing the Arduino data, and the text file containing video data, and verify that you have detected an equal number of stimulus in both files. What can you conclude from these two pieces of data?
• Optional: Open the raw video file and find the exact frame where the stimulus came on. If you compare different trials you might notice that the brightness of the LED in that first frame across two different trials is different. Why is that?

Exercise 5: Trigger a visual stimulus using a button

To make our task more interesting, we will now trigger the stimulus manually using a button press and learn more about SelectMany along the way!

• Connect a new push button component into one of the Arduino digital inputs.
• Insert a DigitalInput source and set its Pin property to the Arduino pin where the new button is connected.
• Configure the PortName to the Arduino port where the microcontroller is connected.
• Insert a Condition operator.
• Insert a SelectMany operator and move the stimulus generation logic inside the nested node:

Why do we need to remove the Repeat operator?

• Ask a friend to test your reaction time!
• Optional: In the current workflow, what happens if you press the stimulus button twice in succession? Can you fix the current behaviour by using one of the higher-order operators?

Exercise 6: Recording response-triggered videos

• Starting from the previous workflow, insert another AnalogInput source with the Pin property set to the button press pin number.
• Insert a GreaterThan operator.
• Insert a DistinctUntilChanged operator.
• Insert a Condition operator.
• In a new branch coming off the VideoWriter, insert a Delay operator.
• Set the DueTime property of the Delay operator to 1 second.
• Insert a TriggeredWindow operator, and set its Count property to 100.
• Insert a SelectMany operator and inside the nested node create the below workflow:

• Run the workflow and record a few videos triggered on the button press.
• Inspect the videos frame by frame and check whether the response LED comes ON at exactly the same frame number across different trials.
• If it does not, why would this happen? And how would you fix it?