Knowledge Base

Data Analysis

Data Preparation

Track Editing

Choose Acquisition > Edit Tracks.

You can fix tracking errors but normally you will not need to edit your data.

Track Smoothing

Choose Acquisition > Track Smoothing Profile > open MDM 0.2 cm. In this profile, the Minimal Distance Moved filter is used (option set to Direct, with a threshold of 2 mm). This filter filters out small movements which result from system noise.

Note that smoothing influences the track, that is, the coordinates of the center point of the subject, not the subject's detected area, and therefore does not influence the calculations of Mobility. It also has no effect on Activity analysis.

Review Video and Behaviors

If you have scored behaviors manually, you can review the video and if necessary edit the data. Choose Acquisition > Score behaviors manually. For details, see Acquire Data > Score behaviors manually in the EthoVision XT Help.

Selecting Data

Choose Analysis > Data Profile.

You can select your tracks according to your independent variable values and also select parts of tracks (nesting). You can also create groups of tracks to obtain group statistics for each group. Figure 9.7 shows an example of a Data Profile to compare the animals that were treated with a drug, with those that were treated with saline.

Integrated Visualization

Choose Analysis > Results > Plot Integrated Data.

Select a trial from the list on the tool bar. The video file and the Time Event Plot are displayed.

In the Time Event Plot tool bar you can choose which profiles to use for display of the variables. Figure 9.8 shows how variables are visualized in the sample experiment Porsolt Forced Swim Test XT180 with the Analysis Profile Mobility_10 containing the variable Mobility.

Play the video and check whether the states scored under Mobility state correspond to the behavior of the animal. If they don't, fine tune the Mobility (or Activity) thresholds in the Analysis profile. Then, check the results of your changes again in the Integrated Visualization.

Analysis Profile

Mobility Analysis

In an Analysis Profile, you select the dependent variables and statistics for analysis. For the experiment made with the template for the Porsolt Test, a dependent variable Mobility is already present.

Choose Analysis > Analysis Profile > Open > Mobility.

1. Double-click the row for the Mobility state variable.
2. Under Outlier filter, the Averaging interval is the number of samples on which the running average mobility is based. With an Averaging interval of 1 sample, sudden changes in surface area caused by reflections are not filtered out.
3. Under Threshold, set the Highly mobile and Immobile thresholds to the values you have previously determined (see determining mobility thresholds on page 171).
4. Under Calculate statistics for, select the states you want to analyze.
5. Click the Trial Statistics tab and select Duration and Frequency. Optionally select additional statistics in the Group statistics tab.
6. Click OK.

Comparing EthoVision Mobility Data with Your Ratings

If you have scored the animal's behavior manually, you can easily compare your data with those scored by EthoVision (Mobility). Make sure that your analysis profile includes the three Mobility states and the behaviors scored manually, then visualize the data.

If the EthoVision data do not match your ratings at an acceptable level, adjust the Mobility threshold values in acquisition and in analysis.

An Alternative to Mobility Analysis

You can use Activity state if the Mobility state variable does not give good results, for instance when detecting thrashing of the animal's limbs which may not be detected properly by Mobility state.

Activity analysis is based on the temporal change in pixel value in the whole arena. Because it is not based on the detected animal's surface area, it is independent of tracking.

1. In the Experiment Settings, under Analysis options select Activity analysis. Click the video button next to it and watch the video tutorial to learn about Activity analysis.
2. In the Detection Settings, click Activity. Enter the Activity threshold for detecting a change in the pixels in the arena. Start with the default value and check the purple-colored pixels in the video image. Adjust the threshold in such a way that when the animal is immobile, those pixels are almost completely absent.
3. Run the trial.
4. In the Analysis profile add Activity and Activity state.
5. Visualize the data (Plot Integrated Data). There you see the scored Activity states and the subject in the video. Adjust the Activity state settings in such a way that the states scored by EthoVision XT match the behavior of the subject at that time.

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