ABSTRACT
The algorithm for such a system is schematized in Figure 2A-F. Once the user sets up the training parameters for all of the dishes, there is a repeated cycle (providing a consistent environment for
Ptir posr cemented into side
some number of hours or days). This cycle consists of image processing to detect the position of each worm (Fig. 2C), decisions (determined by the user-specified learning task) as to whether each worm is to be punished or rewarded and a corresponding change in the light and shock environment of the dish. All of the data (coordinates of the worms’ movements plus real-time movies of their activities) are recorded to disk and may be made instandy available via the Internet. In simpler trials (with no rewards or punishments), the system can simply gather quantitative data on the baseline behavior of control or drug-treated worms and their response to light, shock and vibration.
