Meeting Abstract
The kinematic parameters of crawling behavior in the caterpillar Manduca sexta are being studied to better understand the neuromechanics of soft bodied animal locomotion. We have developed a method to track and analyze the 3D movements of Manduca in real-time, while crawling on different substrates and in different orientations. This method is being used to determine which substrate parameters are sensed by the animal, and how it uses this sensory information to adjust its gait. Manduca movements are complex and variable, so real-time tracking allows detailed kinematic parameters to be analyzed to identify important statistical changes in large datasets. The method is based on a commercial device (Vicon), originally designed to capture human motion, and it reports the three-dimensional position of markers in the scene. To adapt this system for use on small (and soft) animals we have replaced the retroreflective markers with microscopic IR-LEDs. These are manually soldered to one another with a very fine wire to form a suit that is stitched onto the cuticle of the caterpillar. The caterpillar is then placed inside a circular treadmill that can be positioned in any orientation with respect to gravity and the Vicon cameras report the 3-D position of the LEDs. A custom-made, multi-layer parallel-processing Matlab software, working across several computers, acquires the Vicon data and analyzes it in realtime while also saving it for future, offline replay. With this system it is possible to quantify the movements of multiple body segments simultaneously and to detect small changes in the coordination of the crawling gait in response to environmental stimuli. Supported by NSF/IOS grant 1456471 awarded to Barry Trimmer.
