Computer assisted video analysis of swimming performance in a forced swim test: simultaneous assessment of duration of immobility and swimming style in mice selected for high and low swim-stress induced analgesia.

In behavioral pharmacology, two problems are encountered when quantifying animal behavior: 1) reproducibility of the results across laboratories, especially in the case of manual scoring of animal behavior; 2) presence of different behavioral idiosyncrasies, common in genetically different animals, that mask or mimic the effects of the experimental treatments. This study aimed to develop an automated method enabling simultaneous assessment of the duration of immobility in mice and the depth of body submersion during swimming by means of computer assisted video analysis system (EthoVision from Noldus). We tested and compared parameters of immobility based either on the speed of an object (animal) movement or based on the percentage change in the object's area between the consecutive video frames. We also examined the effects of an erosion-dilation filtering procedure on the results obtained with both parameters of immobility. Finally, we proposed an automated method enabling assessment of depth of body submersion that reflects swimming performance. It was found that both parameters of immobility were sensitive to the effect of an antidepressant, desipramine, and that they yielded similar results when applied to mice that are good swimmers. The speed parameter was, however, more sensitive and more reliable because it depended less on random noise of the video image. Also, it was established that applying the erosion-dilation filtering procedure increased the reliability of both parameters of immobility. In case of mice that were poor swimmers, the assessed duration of immobility differed depending on a chosen parameter, thus resulting in the presence or lack of differences between two lines of mice that differed in swimming performance. These results substantiate the need for assessing swimming performance when the duration of immobility in the FST is compared in lines that differ in their swimming "styles". Testing swimming performance can also be important in the studies investigating the effects of swim stress on other behavioral or physiological parameters because poor swimming abilities displayed by some lines can increase severity of swim stress, masking the between-line differences or the main treatment effects.