Meeting Abstract
XROMM studies of mammalian feeding have produced a daunting amount of kinematic data. From temporomandibular joint kinematics to individual tooth cusp trajectories, there are a myriad of ways to quantify the movements that occur during feeding in mammals. To facilitate and standardize post-processing of XROMM data, consisting of synchronized movements of craniomandibular joint coordinate systems (JCS; i.e., rotations and translations) and locators, we developed FeedCycle, a metadata management and data analysis program. FeedCycle uses the gape cycle, determined from Rz (jaw open/close), and intra-cycle phases, determined from the second derivative (acceleration) of Rz, as the basis of analysis of all JCS and locator data. In the primary panel for identifying cycles and phases, users can adjust a threshold and smooth waves with a Butterworth filter to accurately detect cycles. A separate panel showing synchronized graphs of Rz and Ry with imported XROMM, fluoro or light videos allows annotation and manual event marking of individual cycles. Users can move by frame or cycle to verify or override auto-detected parameters such as chewing side, based on Ry, as well as annotate cycles with behavioral data. For each degree of freedom and locator, the following are exported for each cycle and phase along with corresponding metadata as .csv files: frame #, time, duration, maximum, minimum, start and end values. Using examples from 1 and 2 JCSs from our datasets, we demonstrate that FeedCycle accurately and rapidly standardizes data extraction based on fundamental features of mammalian feeding cycles in order to prepare data for subsequent statistical analysis.
