Long-term Behavioral Tracking of Freely Swimming Weakly Electric Fish

The overall goal of this procedure is to record spontaneous spatial exploration by synchronously recording physiological signals and videos from freely swimming fish. This is achieved by constructing a large experimental tank surrounded by a sensory isolation chamber. Next, a multi-channel electrical recording device is constructed to precisely and reliably measure the electric organ discharge.

An infrared video camera is used to synchronously track the animal's body posture and trajectories. Results are obtained that show interplay between an animal's body movements and its active electro sensory sampling rate. The main advantage of this technique over existing methods like eliciting evoked responses by delivering external stimuli to restrain animals is that the spontaneous and naturalistic behaviors from a freely behaving animal can be non-invasively observed.

This method can help answer key questions in the neuropathology field, such as the role of active sensing during self-guided exploratory behaviors. We first had the idea of this method when we began to investigate the role of active electros sensory sampling in pulse type weekly electric fish during self-guided exploration and spatial learning. Visual demonstration of this method is critical due to difficulties with movement induced recording artifacts and the complications associated with lighting and imaging through water.

The chamber rests on an anti-vibration surface containing an electrically shielded heating element over thermally graded foam padding. The aquarium tank consists of a central arena and four corner compartments. The compartments are separated by motorized gates to track electric organ discharge.

Install eight graphite electrodes spaced equally on the curved wall of the central arena. Use cable assemblies to wire all electrodes to the amplifier unit. Differentially amplify by pairing two electrodes oriented 90 degrees from each other.

Make sure to ground all coaxial shielding wires by connecting them to the Faraday cage. Use spike two software to adjust signal processing and add a DC remove process to all recording channels. Next, add a rectify process to all recording channels.

Create a virtual channel by summing all four recording channels. Extract a unimodal envelope per EOD pulse by adding a root mean squared process to the virtual channel. Next, create a real mark channel from the virtual channel.

Set the threshold to a level that captures all EOD pulses while avoiding false positives. To monitor the EOD rate in real time, set the channel display option of the real mark channel to an instantaneous frequency mode. Fish movement is monitored in real time by duplicating the real mark channel and setting the display option to a wave four mode.

To calculate the activity level from the root means square of the EOD amplitude slope, create a virtual channel from the real mark channel and then add slope and RMS processes. Export the real mark channel in the Spike two software to the mat lab format. To create a background scene, cover any object that casts a reflection on the water surface with matte white countertop film.

Install a matte white panel below the ceiling to hide the camera and air vent. Next, install white LED lights for the maintenance of the diurnal light cycle and infrared LED lights for imaging in darkness. Direct all light sources toward the ceiling to achieve indirect and uniform illumination.

Point a near infrared camera through a small viewing hole in the ceiling panel, and make sure the viewing angle is wide enough to image the whole central arena. To make a time synchronized video recording, place an infrared LED at one of the four tank corners. To generate time synchronization pulses.

Add a load limiting resistor in series and drive the LED from a digital output port of the digitizer hardware. Using specialized software, select the highest recording quality and the highest resolution supported. Start the video recording immediately before starting the EOD recording.

After placing the fish in the chamber, record spontaneous behavior for a predetermined period of time. After the recording, convert the image numbers to the Digitizer time unit by linearly interpolating between the first and the last light pulses captured by the signal digitizer and the video recording. Using the included software based on matlab, import a video recording file Using the video reader read function.

Create a composite background image by combining two image frames. Replace the image region occupied by an animal with an unoccupied image of the same region from another frame. Specify an image region to track by drawing a circular mask around the central arena.

Multiply by a constant to set a minimum threshold for intensity difference. Next, subtract an image frame from the background image to obtain the difference image. After applying an intensity threshold, cleaning the binary image and calculating the blob areas, select the largest blob corresponding to an animal.

Next, rotate the image to align the major axis with the X axis and divide the image into head and tail parts. At the oid, identify the major axis of the head part and rotate the entire image to align with the x axis. Fit bounding boxes around the head and tail parts parallel to their major axis.

Next, determine the median why coordinates of the blob at the left center and right vertical edges of the bounding boxes, and assign them to five feature points. Locate the animal cent crop and image frame centered at this point and continue to process successive frames. Manually assign the head orientation for the first frame and use a dot product between the orientation vectors from two successive frames to automatically determine the head orientation.

Inspect the result and manually flip the head orientation if incorrectly assigned. Plot an animal trajectory by joining the head tips and smooth with filters if it has a jittery appearance. After super imposing the trajectory with a background image, interpolate the fish midlines.

Using the five feature points, compute the average EOD rate at each image capture time by resampling and averaging the instantaneous EOD rate. After plotting the trajectory in pseudo colors determined from the time matched EOD rate, the images superimposed with the background. The recorded EOD wave forms from different electrode pairs varied in amplitudes and shapes as expected from their unique positions and orientations.

The successive EOD peaks and instantaneous EOD rates were joined and linearly interpolated at constant time interval. The EOD amplitudes recorded at external electrodes remained constant while an animal was at rest, but varied over time while the animal moved. While the fish was at rest.

The baseline EOD rate remained low, but it became significantly higher while the fish actively swam. The animal's trajectory and midlines are shown here during an abrupt turning behavior. In this image, a pseudocolor representation of the average EOD rate is superimposed with the time matched trajectory of the fish's head tip.

This example shows the visual tracking output generated by the automated image tracking procedures described previously. While attempting this procedure, it's important to remember to minimize external background noise, which could influence spontaneous animal behaviors. The development of this technique paves the way for researchers in the field of neuropathology to explore active sensing during self-guided movements in the presence of complex obstacles.

After watching this video, you should have a good understanding of how to construct the experimental tank and set up data acquisition to track spontaneous behaviors for long durations. Don't forget that working with high voltage equipment near a wet environment can be extremely hazardous, and the precautions such as keeping the tank surroundings dry and wearing insulating gloves should always be taken while performing this procedure.