SEE ("Software for the Exploration of Exploration") is an advanced program for the visualization and analysis of rodent spatial behavior in the open field [1]. Rather than employing arbitrary, ad-hoc parameters introduced by the investigator or the automated measurement system (e.g. "activity"), SEE displays and quantifies the intrinsic patterns or "units" of which spatial exploratory behavior consists. The first and necessary stage of SEE analysis is the segmentation of the animal's path into the most basic of these intrinsic units: stops (within place behavior) and movement segments (going between places). This segmentation is based on the empirical distribution of movement speeds, in which more than one component can typically be shown to exist [2]. It was essential therefore to determine how general these components are, and how independent they are of arena size, experiment treatment and tracking system properties.

SEE was developed and demonstrated with data of the normal behavior of rats and mice, measured by a video tracking system in large (3 m to 6.5 m) circular arenas. Here we report its application to data from drug-injected rats, measured by a standard photo-beam chamber 43 cm wide at a much lower rate. Subjects were Sprague-Dawley rats, injected with phencyclidine (five doses including saline) or d-amphetamine (four doses including saline). The raw data of the animal's coordinates were fed to the same SEE segmentation algorithm used with the video tracking data. Results show that, with both drugs and with all doses, components of stops and movement segments are statistically significant in the distribution of speeds. Typical values of the high-speed component were, as expected, lower relative to those found in large arenas. As in the large arena, however, there was a significant difference between the spatial spread of the components, typically 5-10 cm for the low-speed component and 20-30 cm for the high-speed component, indicating that they indeed represent different behavioral categories of moving within a place versus going between places. There was no need for any substantial altering of the algorithm in order to accommodate for the very different experimental conditions.

These results support the generality of the intrinsic categorization of rodent's spatial behavior into stops and movement segments. They also suggest that the full power of SEE analysis can be applied to photo-beam measurement of drug-induced locomotor behavior. Several possibilities for sophisticated querying, visualization and quantification of such behavior will be demonstrated.

Researchers interested in using SEE are encouraged to contact Ilan Golani (ilan99@post.tau.ac.il).

This research was supported by a grant from Novartis.

References

1. Drai, D.; Elmer, G.; Benjamini, Y.; Kafkafi, I.; Golani, I. (2000). SEE: software for the exploration of exploration. This volume.
2. Drai, D.; Benjamini, Y.; Golani, I. (2000). Statistical discrimination of natural modes of motion in rat exploratory behavior. Journal of Neuroscience Methods, 96, 119-131.

© 2000 Noldus Information Technology b.v.