Knowledge Base

Distance Moved

Definition

The distance traveled by the center, nose or tail-base point of the subject from the previous sample to the current one. It is calculated as:

• DMk = Distance moved from sample k–1 to sample k
• Xk-1, Yk-1 = X,Y coordinates of the center, nose or tail-base point at sample k-1
• Xk, Yk = X,Y coordinates of the center, nose or tail-base point at sample k-1

How to Specify Distance Moved

1. Click Add next to Distance moved.
2. Complete the procedure to add the variable. See Calculate statistics: procedure.

Notes

• If your experiment is set to Center-point, nose-point and tail-base detection, click the Body points tab and select the body points for which you want to calculate the distance.
• Because it is based on change in X,Y coordinates, Distance moved needs two valid (non-missing) samples. A missing sample in your track results in missing Distance moved values in that and in the next sample. Make sure that the proportion of missing samples is low (less than 1%).
• Sample rate influences the values of Distance moved. When tracking at too low a sample rate, parts of the actual path are cut off, resulting in an underestimation of per-sample (and total) Distance moved. If, on the other hand, the sample rate is too high, EthoVision XT catches the wobbling of the body's center point of the walking animal, causing extra apparent movement, and therefore an overestimation of per-sample (and total) Distance moved. See Track Smoothing for a description of how to filter out small movements.

Applications

Distance moved is often used to give a general measure of activity. It is also used as the basis for calculating other parameters such as velocity (see below).

In rodent models of aging, the total and maximum distance moved are used to calculate the frailty index. See Parks et al. (2012) J. Gerontol. A 67(3): 217-227.

Velocity

Definition

The distance moved by the center, nose or tail-base point of the subject per unit time.

Velocity is obtained by dividing Distance moved by the time difference between a sample and the previous one, where Vk = velocity at sample k (expressed in the unit you have defined in the Experiment Settings) and DMk = Distance moved at sample k.

How to Specify Velocity

1. Click the Add button next to Velocity.
2. Under Outlier filter, specify the Averaging interval. Leave 1 if you want to keep the raw data; enter a value to smooth the values using a running average method. See Averaging interval.
3. Complete the procedure to add the variable. See Calculate statistics: procedure.

Notes

• If your experiment is set to Center-point, nose-point and tail-base detection, click the Body points tab and select the body points for which you want to calculate the velocity.
• Sample rate influences the calculation of Distance moved, and therefore Velocity. When tracking at too low a sample rate, parts of the actual path are cut off, resulting in an underestimation of per-sample Velocity. If, on the other hand, the sample rate is too high, EthoVision XT catches all random movements and wobbling of the body point of the walking subject, causing extra apparent displacement, therefore an overestimation of Velocity. See Track Smoothing for a description of how to filter out small movements.
• Because it is based on change in distance moved, Velocity needs two valid (non-missing) samples. A missing sample in your track results in missing Velocity values in that and in the next sample. Make sure that the proportion of missing samples is low (less than 1%). You can view the proportion of missing samples as one of the System Variables in the Trial list.

Applications

• Apart from the obvious applications of this dependent variable, the mean velocity is sometimes used as a measure of general activity (for example, Nilsson et al. 1993, J. Exp. Biol., 153-162; Winberg et al. 1993, J. Exp. Biol., 213-232).
• TIP: Define Velocity and Movement in the same Analysis profile to find out which averaging interval results in the best match between movement bouts and video.
• In rodent models of aging, the average velocity is used to calculate the frailty index. See Parks et al. (2012) J. Gerontol. A 67(3): 217-227.

Averaging Interval

Aim

To smooth the values of a dependent variable.

This Topic Applies To

Dependent variables: Acceleration state, Body angle, Activity state, Body angle state, Body elongation, Body elongation state, External data (resampled), External data - state, Rotation, Mobility state, Movement, Velocity.

The dependent variable can be in a condition defined in the Trial Control Settings, in the Analysis profile and in the Data profile. Setting the Averaging interval in one part of EthoVision XT does not influence the value of the same variable in the others.

How to Access This Option

In the Trial Control Settings, in the Analysis profile or in the Data profile select the dependent variable and locate Outlier filter.

How Averaging Interval Works

• When Averaging interval is 1, the outlier filter is off, thus the values of the variable are not smoothed.
• When Averaging interval is 2 or larger, EthoVision XT replaces the per-sample value of the dependent variable with the average calculated over the number of samples specified by the interval.

The table below shows how EthoVision XT re-calculates a variable V in a few samples when Averaging interval is set to 2. Note how the average (avg) is obtained when values of the dependent variable are missing. Remember that velocity V at time t is only calculated when there are valid samples at time t and t-1.

Notes

• Note the difference between Track Smoothing and the Outlier filter:
  • With Track Smoothing, you smooth the raw x,y coordinates. This also has an effect on the dependent variables calculated based on those coordinates, for example Distance moved. See Smooth the Tracks.
  • With Outlier filter, you smooth the values of the dependent variable, for example velocity or mobility, after they are calculated from the raw data. The Outlier filter is useful when you want to calculate state variables. Apply the Outlier filter, for example, when you want to smooth Velocity to calculate the Movement states, which are based on velocity; or smooth Mobility when you want to calculate Mobility state.
• If you combine Track Smoothing with Outlier filter, the dependent variable is calculated with the Outlier filter after the raw coordinates are smoothed with Track Smoothing.

Source: EthoVision XT 19 Help, Noldus Information Technology