Knowledge Base

Conditions Based on Dependent Variables

Selecting Subjects for a Condition

1. If the Actors tab also contains the Body points options, select one or more body points of the subjects selected above.
2. For the dependent variables of social interaction, a Receivers tab is also available. Click this tab and select the subjects that you have not selected under Actors.
3. Click OK to confirm your selection. You return to the condition window.

Example: Using All Selected Subjects

The researcher wants to define an In zone condition that is met when Subject 1 and Subject 2 have been in a zone simultaneously for at least one minute.

Solution: In the Condition window, select Cumulative duration as Statistic and enter >= 60 s. Click Settings and select the zone you require, then click Actors and select Subject 1 and Subject 2, finally select All selected Subjects.

Example: Using Any Selected Subjects

The researcher wants to define a Distance to zone condition that is met when one of the two subjects is at less than 5 cm from a novel object.

Solution: In the Condition window, select Current and enter <= 5 cm. Click Settings and select the novel object zone, then click Actors and select Subject 1 and Subject 2, finally select Any selected Subjects.

Example: Using Actors and Receivers

The researcher wants to define a condition that is met when the nose point of Subject 1 has been close to at least one body point of Subject 2 for more than one minute.

Solution: Click the button next to Proximity and in the Condition window, select Cumulative duration and enter the required value. Click Settings, then:

• In the Actors tab, select Subject 1 and de-select Subject 2. Under Body points, select Nose-point only.
• In the Receivers tab, select Subject 2 and de-select Subject 1. Under Body points, select all points and then Any selected points, since the condition must be met when Subject 1 is in proximity of at least one of Subject 2's body points.

When Is a Condition Met?

Dependent variables are calculated for each sample. When a condition based on dependent variables is met depends on a number of factors.

Effect of Outliers and Track Smoothing

Dependent variables used in conditions are not smoothed or filtered using the Lowess filter or the Minimum distance moved as they can be for data analysis. This means that outliers, nose-tail swaps, or even subject identity swaps can have severe effects on whether and when a condition is met.

For example, if EthoVision XT detects a reflection within zone A as the subject, the track shows a spike. If a condition of type In zone is defined which specifies Frequency of In zone A >= 1, then the condition is met the moment the subject is wrongly detected in zone A, assuming that the condition box was active.

Track noise reduction may solve this problem. In the Detection settings, click Advanced and under Smoothing set Track noise reduction to On. This applies an online smoothing effect as the subject is being tracked.

Note: This change in the track cannot be reversed afterwards.

Effect of Missing Samples

Missing samples can also cause Trial Control to react later. For example, an In zone condition is met only when valid samples are found within the specified zone.

If EthoVision XT misses the subject some time before it enters zone A, the following applies:

• If a condition of type In zone specifies Current of In zone A = true, then the condition is met later than it would be if there were no missing samples.
• If a condition of type In zone states "Current duration of not in zone longer than 1 s" (assuming a sample rate of 5/s, so 5 samples make up 1 second), then the condition does not become true, because there are only 4 samples for which the state "not in zone" was true.
• The condition "Current value of In zone A = false" may be met in the first segment of the track, because it refers to the state of a current sample, not a duration. In that case the subject is out of the zone, so the current value of In zone is "false". However, the condition will not be met in the period of missing samples because EthoVision XT does not know the current state of the subject.

Effect of the Statistics Used

When a condition is evaluated also depends on the statistic chosen:

• For Current, the condition is evaluated at the same sample time as when it is activated.
• For all other statistics (such as Frequency, Current duration, etc.), it is evaluated at least one sample after the condition is activated. This depends on when the first value available for comparison occurs after the activation of the condition.
• When using Velocity and Distance, it takes one sample.
• When using Heading and In zone, it takes two samples.
• When using Turn Angle and Angular velocity, it takes three samples.

Conditions Based on Behaviors Scored Manually

Use this kind of condition to have EthoVision XT carry out an action when the subject performed a behavior you scored manually. For example, in a social interaction test, to stop tracking when the subject attacked three times, or when the duration of Intimidation lasted more than 30 seconds.

1. In the Components pane, under Conditions, double-click the name or click the button next to the manually scored behavior you want to base the condition on. If you do not see the variable name, click the More button.
2. The Manual Scoring Condition window opens. Under Name, enter a Condition Name. Under Comment (at the bottom) enter a description (optional).
3. Under Condition is met when, click the Settings button. If the behavior is of the type Start-Stop or Mutually Exclusive, choose the behavior to base the condition on, for example Rear or Not Rear.
4. In the Actors tab, select the subjects to which the condition applies. If the Actors tab also contains the Body points options, select one or more body points of the subjects selected above. Click OK.
5. Under Condition is met when, select the Statistic you want to use.

Common cases of Statistic for manually scored behaviors:

• Current. For a condition that is met when the behavior at the current sample is true or not true. This statistic is available for Start-Stop and Mutually exclusive events.
• Frequency. For a condition that is met when the behavior occurred a number of times.
• Cumulative duration. For a condition that is met when the animal has shown the behavior for a specific time since the activation of the condition. This statistic is available for Start-Stop and Mutually exclusive events.
• Current duration. For a condition that is met when the animal has shown the behavior continuously for a specific time. This statistic is available for Start-Stop and Mutually exclusive events.
• Latency to first. See the Latency to First section for details.

For more information on manually scored behaviors, see the EthoVision XT Help.

Conditions Based on Hardware Devices

Use this kind of condition to check the status of hardware devices (pellet dispensers, lickometers, etc.) and have EthoVision XT take actions when the conditions are met.

1. In the Components pane, under Conditions, click the button next to the hardware device you want to base the condition on.
2. The Hardware condition window appears. Under Name, enter a Condition Name. Under Comment (at the bottom) enter a description (optional).
3. Under Condition is met when, select the Template device you want to use. A template device represents a group of devices of the same type, one for each arena.
4. Under Signal to check, select the criterion that forms the condition.
5. Under Signal value, select the value of the criterion in order for the condition to be met.
6. Click OK. Insert the Condition box in the appropriate position in the rule.

Hardware devices are not defined for each arena separately. If you have a setup with multiple arenas, it is assumed that each arena has the same type of device. The name in the device list is the generic name valid for all arenas. The condition will be evaluated for each arena separately.

If the device type is not listed under Conditions, it means that it is not yet recognized by EthoVision XT.

Pellet Dispenser

• Number of drops. To base your condition on whether more or less than N food items have been delivered since the activation of the condition. Choose >= or <= and enter the value N in the box.
• In error state. To base your condition on whether the pellet dispenser is in error state, that is, when it can no longer drop any pellet (this happens when there are no pellets left in the feeder). Choose between true (error) or false.

Example: Have EthoVision XT give a sound signal when the feeder has run out of pellets.

Lickometer

• Duration of licks. To base your condition on the total duration of licks since the activation of the condition. Enter this time in the box.
• Number of licks. To base your condition on the number of licks since the activation of the condition. You can select >= N, <= N, or = N.

Top Unit of PhenoTyper

• Choose the option under Signal to check to base your condition on whether a sound or light signal is given since the activation of the condition.
• For Top Unit (NeuroBsik), use Is Fsel-1/2 on/off or Is Vol-1/2 on/off to base your condition on the value of frequency or volume specified by one input. Since the frequency and volume are specified by two inputs, you must define two conditions, one for input 1 and one for input 2. Combine the two conditions with an operator of type All.

Note: Neuro-Bsik is a consortium of eleven Dutch Neuroscience research groups and two companies including Noldus. The consortium aims to contribute to the battle against brain disorders by developing novel mouse models for brain disorders. For this project, Noldus has developed a special Top Unit of the PhenoTyper cage that can produce ultrasound stimuli. For more information, see the PhenoTyper - EthoVision XT Reference Manual.

TTL Port Tester

• Button 1/2 presses. To base your condition on the number of presses of button 1 or 2 given since the activation of the condition. You can select either >= N, <= N, or = N. Enter the value N in the box.
• Is button 1/2 pressed? To base your condition on whether button 1 or 2 is currently pressed. Select true or false from the list.
• Is light 1/2 on? To base your condition on whether light 1 or 2 is currently on. Select true or false from the list.

Beam Controller

• Beam 1/2 breaks. To base your condition on the number of beam breaks for beam 1 or 2 since the activation of the condition. You can select either >= N, <= N, or = N. Enter the value N in the box.
• Is Beam 1/2 broken? To base your condition on whether beam 1 or 2 is currently broken. Select True or False from the Signal value list.
• Is cue 1/2 on? To base your condition on whether cue 1 or 2 is currently on. Select true or false from the Signal value list.

Activity Wheel Counter

• Quarter cycles. To base your condition on the number of quarter cycles since the activation of the condition. You can select either >= N, <= N, or = N. Enter the value N in the box.

Custom Hardware Device

This includes the illuminated shelter.

• Input 1/2 L/H -> H/L count. To base your condition on how many times there has been a transition from High to Low (or vice versa) in input 1 or 2 during the time t-1 to t, where t is the sample time when the condition is evaluated. Choose the option that applies, select either >= N or <= N. Enter the value N in the box. This option lets you create a condition based on events (transitions) in the signal, no matter when they happen during the sample time.
• Is input 1/2 High/Low? To base your condition on whether the current state of input 1 or 2 is High or Low at the sample time when the condition is evaluated. Select the option that applies, and then either true or false.
• Is output 1/2 High? To base your condition on whether the current state of output 1 or 2 is High at the sample time when the condition is evaluated. Select the option that applies, and then either true or false. This option evaluates the state of the signal (High vs Low) at the sample time when the condition is evaluated. If a quick transition High -> Low -> High occurs within the sample interval, this is not noticed; the state at the sample time is still High.

Counts vs. Signals: Effects on Conditions

Consider two conditions based on the same input (for example Input 1):

• Is the number of transitions H->L >= 1?
• Is the state of the signal Low?

There is a fundamental difference between the two conditions:

• In the first condition, the software counts the number of transitions from High to Low during the sampling interval (that is, from t-1 to t). The condition may become true or stay false for sample t, depending on that value.
• In the second condition, the signal is evaluated at time t, that is, at the end of the sampling interval. If short transitions occur from High to Low to High again, the Low signal is not detected by this condition, because the signal is still High at time t.

More About Conditions Based on Dependent Variables

This section provides details and examples of the statistics you can use for your condition.

Note: Dependent variables of Live Mouse Tracker and the social variables Train and Side by side are not available in the Trial Control Settings.

Distance Moved

Frequently-used statistics:

• Current (default). To base your condition on the distance between the previous sample and the current one. This value depends on the chosen sample rate. Use Current velocity instead.
• Total. To base your condition on the distance traveled from the moment the condition becomes active up to that point.
• Mean. To base your condition on the average distance traveled from one sample to the next, calculated from the activation of the condition box up to that time.

Source: EthoVision XT 19 - Trial and Hardware Control Reference Manual