Mobility
The percentage changed pixels of the detected subject between current sample and previous sample.
note Mobility is not available if your experiment is set to:
▪Live Mouse Tracker (but see Stretch attend posture).
▪Center-point, nose-point and tail-base detection, when you track two subjects per arena with Deep learning.
Step 1 - Calculation of the changed area
All pixel coordinates of the subjects are determined immediately after they have been detected. Those coordinates are compared with the previous sample to determine the number of changed pixels between the two. The changed pixels are
▪The subject pixels found in the current sample but not in the previous sample AND.
▪The subject pixels found for the previous sample but not for the current sample.
This can be expressed in the following formula:
CAk = (Ak - Ak-1) + (Ak-1 - Ak)
Where CAk is the changed area for the current sample k, Ak is the area for the sample k, and Ak-1 the area for the sample k–1.
In the example below, CA is the sum of the orange areas when comparing the contour of the subject with that at the previous sample. The sum equals the symmetric difference between sets of points known in mathematics and is generally indicated with delta (D). The changed area is shown for samples 2 and 3.
The changed area (in number of pixels) is available when you export your raw data. Open the export file and locate the column named Areachange. Note that Areachange is not the same as the difference in Area between the current and the previous sample!
Mobility is calculated by taking every pixel identified as the subject and comparing it between the current image and the previous one. Note that Mobility also includes movement in space (e.g. walking).
Step 2 - Calculation of Mobility
The formula for Mobility is the changed area for the current sample k divided by the sum of the current area and the previous area:
To smooth the Mobility parameter, an Averaging interval is used. This gives you the option to specify the number of samples for calculating a running average of Mobility. The Mobility percentage calculated as above is averaged over the number of samples that you specify. This way, sudden changes in surface area caused by such factors as the animal entering a shadowed area and not being identified correctly, or a reflection being identified momentarily as the animal, are smoothed out. See also Averaging interval
When you export data, you do not export any information about which averaging interval was used.
Range
Mobility ranges from 0% to 100%.
▪If all the pixels of the detected subject are equal in value between sample k-1 and sample k, for example when the subject is completely still, Mobility at sample k is 0.
▪If the animal moves and increases its velocity, whilst keeping the same shape, there will be an increase in mobility, because the pixels belonging to the animal are increasingly different as it moves faster.
▪when the subject is so fast that its contour at consecutive samples does not overlap anymore, then all the pixels are different. CA is maximal and equals the sum of the two contours. Therefore, Mobility is 100%.
1.Open an Analysis profile and in the Dependent Variables panel, under Body, click the Add button next to Mobility.
2.Select the Averaging interval. See above Step 3 - Running average for more information on the Averaging interval.
3.Click the Trial Statistics tab and select the statistics you want to calculate.
Notes
▪When the experiment is set to Color marker tracking, that is, when EthoVision only tracks the markers and ignore the body contour of the subject, Mobility is calculated based on the surface area of the marker.
See also