1. Anatomically-based skeleton kinetics and pose estimation in freely-moving rodents
- Author
-
Jakob H. Macke, Kay-Michael Voit, Jason N. D. Kerr, Edyta Leks, Jürgen Sawinski, Klaus Scheffler, Damian J. Wallace, and Arne Monsees
- Subjects
Neural activity ,Gait (human) ,business.industry ,Computer science ,Kinetics ,Biomechanics ,Computer vision ,Artificial intelligence ,business ,Videography ,Skeleton (computer programming) ,Pose ,Surface anatomy - Abstract
Forming a complete picture of the relationship between neural activity and body kinetics requires quantification of skeletal joint biomechanics during behavior. However, without detailed knowledge of the underlying skeletal motion, inferring joint kinetics from surface tracking approaches is difficult, especially for animals where the relationship between surface anatomy and skeleton changes during motion. Here we developed a videography-based method enabling detailed three-dimensional kinetic quantification of an anatomically defined skeleton in untethered freely-behaving animals. This skeleton-based model has been constrained by anatomical principles and joint motion limits and provided skeletal pose estimates for a range of rodent sizes, even when limbs were occluded. Model-inferred joint kinetics for both gait and gap-crossing behaviors were verified by direct measurement of limb placement, showing that complex decision-making behaviors can be accurately reconstructed at the level of skeletal kinetics using our anatomically constrained model.
- Published
- 2021