Knowledge Base

Create an Experiment

For Live Mouse Tracker

1. Choose File > New and in the dialog that opens select Live Mouse Tracker experiment.
2. For the rest of the procedure, see the EthoVision XT Help.

For All Other Cases

Choose File > New From Template and choose a predefined template in the guided setup. See below.

Experiment Settings

Choose Setup > Experiment Settings.

• Under Video Source, click the video icon in the camera row and adjust the camera settings if necessary.
• Under Subjects, specify the Subject roles (for example Resident or Intruder, or Subject 1 and Subject 2). Do not enter the ID of the individual animals.
• Under Tracked Features, select the option that corresponds to your needs. In most cases that is Center-point, nose-point and tail-based detection.
• Under Body Point Detection Technique, select Contour-based if you use color-marked animals, or Deep learning if your subjects look different (e.g. the back of one individual has been partially shaved).

Note: You can use Deep learning to track two subjects per arena. The Deep learning technique works if you have a recent graphics card and the driver is up to date.

Manual Scoring Settings

Choose Setup > Manual Scoring Settings.

Here you can define behaviors that you score manually, for instance by pressing a keyboard key.

In the template experiment, two start-stop behaviors, Rearing and Grooming, have been defined. Define other behaviors if needed, for example Sniffing and Boxing.

For more information, see Set Up an Experiment > Manual Scoring Settings in the EthoVision XT Help.

Trial Control Settings

Choose Setup > Trial Control Settings > New.

In the Trial Control Settings, define conditions for the start and stop of the track.

• Starting condition: For instance, 'Start tracking 5 seconds after both animals are detected in the arena'. If you are tracking live, this 5-second delay enables you to put both animals in the arena and to take out your hand before tracking starts (5 seconds might not be enough; check beforehand how much time you need).
• Stopping condition: In order to be able to compare the tracks in the analysis, they must have the same length. Either define a Maximum Trial duration or a Stop track condition based on time.

Tip: To make sure that tracking starts when both individuals are in the arena, not before, click Settings in the Condition box before the Start track box. Select Current Duration from the Statistic list, then click Settings and then Actors. Select both subject names and select All selected subjects from the list.

Detection Settings

Choose Setup > Detection Settings > Detection Settings 1.

Important: Follow the instructions in the order described below.

If you selected Deep learning in the Experiment Settings, set the sample rate (below), and skip the next steps. The software is able to recognize the subjects automatically so you do not need to adjust subject contrast, size, contour filters etc.

Sample Rate

Click Video. Select a high sample rate: 25 (for PAL analog cameras) or 30 (for NTSC analog cameras and digital cameras) samples per second.

Automated Setup

Once the animals are in view, click Automated Setup. Select the type of animal you are going to test. Click Next and draw a rectangle around each subject. Make sure that you do this when the animals are not in close contact.

If detection of the subject is not optimal after using the Automated Setup function, see Configure Detection Settings in the EthoVision XT Help for details on the advanced detection settings.

Subject Identification for Color-Marked Subjects

If you mark animals with colors, under Subject Identification make sure that Marker-assisted identification is selected.

1. Put the marked animals in the arena or play the video. Make sure to select a point in the video where the animals do not touch each other. Alternatively, place the animals in the arena one at a time.
2. In the Subject Identification section, select one of the subjects and click the Identification button. As a result, the Identification [Subject name] window opens.
3. Move the mouse pointer to the Video window so the pointer becomes an eyedropper. Move the eyedropper over the color marker of the subject you want to identify and click the left mouse button.
4. Fine-tune the color settings by adjusting the Hue, Saturation and Brightness in the Identification [subject role name] window. Change the range of color settings by changing the numbers or by resizing the Hue box on the vertical color bar, or resizing/moving the box in the color map (horizontally to adjust Saturation, vertically to adjust Brightness). As a result, the outline covers (almost) the complete marker.
5. Next, play the video to see in the Video window whether the marker is detected correctly in different parts of the arena. If the marker 'dances' then your color settings are too sensitive. Go back to step 4 and make the box larger.
6. When you are finished fine-tuning the color settings, continue with setting the Minimal marker size: increase the Minimal marker size until, first, noise is not detected anymore and, second, the marker is not detected anymore (as a result, the outline in the Video window disappears). Now enter a value for the Minimal marker size that is somewhere in between.
7. Click OK when you are done and repeat steps 1–6 for the other subjects.

For details, see Advanced detection settings for color marks in the EthoVision XT Help. Also watch the video Identification Settings in the EthoVision XT video tutorial.

Set the Subjects' Contrast

Make sure you have selected the correct contrast and check the following:

• When you use Static Subtraction or Dynamic Subtraction, move the slider to define the animal's contrast. The animal must be fully detected in all parts of the arena and the noise must be minimal.
• For Dynamic Subtraction, set the Current frame weight slider to a value between 1 and 5 to compensate for the changes in the background with time.
• When you use Differencing, adjust the position of the Sensitivity slider.

Source: EthoVision XT 18 - Application Manual, Noldus Information Technology