Knowledge Base

Select Samples Automatically

Aim

To easily find:

• Missing points.
• Samples separated by a distance greater or smaller than a specific value. For example, to find points lying far from the normal path.
• Samples with a surface area greater or smaller than a specific value. For example, to find all points where the detected blob is too large to be the actual subject.
• Samples with the distance between nose-point and tail-base greater or smaller than a specific value.

To Access This Function

Choose Acquisition > Edit Tracks, and select the trial you want to view.

Procedure

1. On the toolbar, click the Auto Select button.
2. In the Auto Select Samples window, under Select, choose the option you require. For more information, see the notes below.
3. Under Scope, select the arena you want to search. If you want to find samples in all arenas, select All Arenas. Select the subject you want to search. If you want to find samples in all subjects, select All Subjects.
4. Click Select All. The samples matching the criterion are highlighted in the track plot. In the Samples List, the corresponding rows are highlighted in green.
5. If you chose Missing Samples in step 2, click the Next >> and <<Previous buttons to move to the next or previous missing sample, respectively. There, you can decide whether to Interpolate Points or leave the sample as missing.
6. If necessary, change the settings under Select and click the Select All button. When you are satisfied with the selection, click Close.

Options

Missing Samples

Choose this option to select all missing samples in the track, for example to interpolate them.

Sample Distance

EthoVision XT selects all the non-missing points separated from the previous or the next by a distance greater or smaller than the threshold you specify. To select samples, EthoVision XT always uses the distance between center points.

Note: Using > and < does not result in complementary selections. Some samples may be selected in both cases.

Tip: To remove outliers, for example data points resulting from detection of reflections in a water maze, select > and enter a value of distance. This value must be greater than a realistic estimate of the maximal distance that the subject could move between two consecutive samples. Note that this value also depends on the sample rate. The higher the sample rate, the shorter the distance moved by the subject between two consecutive samples. Next, use Delete Points and then Interpolate Points to replace the outliers.

Surface Area

For this option, specify the value in the measurement unit selected in the Experiment Settings.

Nose-Tailbase Length

This is only available if your experiment is set to Center-point, nose-point and tail-base detection. Nose-tailbase length is the total length of the segments between nose-point and center-point, and that between center-point and tail-base point.

Restrict to Samples Visible in Plot

Select this option if you want to select only the samples that match the Select criterion and that are currently displayed in the track plot. The other samples that match the criterion but are not displayed are not selected.

Note: If you select this option and then play back the track, the samples that are not currently visualized but that match the selection criterion under Select are not selected. To have them selected, pause the video and click Select All once again.

To Deselect Samples

Do one of the following:

• Press Esc.
• Click anywhere on the background image that is not a point.
• Right-click anywhere on the background image.

Swap Subjects and Body Points

What Do You Want to Do?

Swap subjects:

• Swap subjects
• Prepare the data in multi-subject trials

Swap body points:

• Swap nose-point and tail-base
• Swap nose-point and tail-base point manually
• Swap nose-point and tail-base point automatically

Swap Subjects

Aim

To exchange the identities of two subjects within one sample (or range of samples, or the entire track), after checking that these were assigned incorrectly by EthoVision XT.

An example of subject identity swap: before the swap, Subject 1 has a yellow track and Subject 2 has a red track. At the moment of the swap, the mouse with the yellow track becomes Subject 2 and the mouse with the red track becomes Subject 1. To correct the subject identity swap, swap the subjects for all samples beginning from the moment of the swap up to a new swap, if present, where subjects are labeled correctly again.

Notes:

• If a large number of samples contain body points assigned to the wrong subject, it may mean that the detection settings defined to discriminate between subjects are not optimal, or that you need to improve the lighting or the color marking of your setup.
• You can click Save on the toolbar so that if you are editing many swaps and make a mistake, you do not lose your previous edits. See Saving the Track Edits.

Procedure

1. If you work with multiple arenas, choose first the arena in the Samples List.
2. Play the track until you find samples with wrongly-assigned subjects. Tip: Play frame-by-frame to position the video when the first samples are swapped.
3. Select the sample (or range of samples) in which you want to swap the subjects. In the Samples List, double-click the time value highlighted in blue, then drag to select the rows corresponding to all the samples to swap. The last row selected should be immediately before the next swap where body points are assigned correctly again. Important: If you drag on the Time column, all subjects in the trial are selected. If you have more than two subjects (for example, three subjects and you want to swap Subject 1 and 3), do not drag in the Time column. Instead, select the cells of one of the two subjects to swap. In the next step you will be asked to select the other subject.
4. Click the Swap Subjects button on the toolbar, or press D.
5. If you have more than two subjects, the Swap Subject window appears. Select Swap Selected Samples, select the second subject of the pair you want to swap, then click Apply.