3D Pose Estimation Enables Virtual Head Fixation in Freely Moving Rats

Abstract

The impact of spontaneous movements on neuronal activity raised the need to quantify behavior. We present a versatile framework to directly capture 3D motion of freely definable body points in a marker-free fashion with high precision and reliability. Combining the tracking with neural recordings revealed multiplexing of information in motor cortex neurons of freely moving rats. By integrating multiple behavioral variables in a model of the neural response, we derived a virtual head fixation, where the influence of specific body movements was removed. This enabled analyses of the behavior of interest, i.e. in our case front paw movements. Thereby, we unveiled an unexpectedly large fraction of neurons in motor cortex with tuning to paw movements, which was previously masked by body posture tuning. Once set up, our framework can be efficiently applied to large data sets while minimizing the experimental work load caused by animal training and manual labeling.