Abstract
Motor skill learning enables organisms to interact effectively with their environment, relying on neural mechanisms that integrate sensory feedback with motor output. While sensory feedback, such as auditory cues linked to motor actions, enhances motor performance in humans, its mechanism of action is poorly understood. Developing a reliable animal model of augmented motor skill learning is crucial to begin dissecting the biological systems that underpin this enhancement. We hypothesized that continuous auditory feedback during a motor task would promote complex motor skill acquisition in mice. We developed a closed-loop system using DeepLabCut for real-time markerless tracking of mouse forepaw movements with high processing speed and low latency. By encoding forepaw movements into auditory tones of different frequencies, mice received continuous auditory feedback during a reaching task requiring vertical displacement of the left forepaw to a target. Adult mice were trained over 4 d with either auditory feedback or no feedback. Mice receiving auditory feedback exhibited significantly enhanced motor skill learning compared with controls. Clustering analysis of reaching trajectories showed that auditory feedback mice established consistent reaching trajectories by Day 2 of motor training. These findings demonstrate that real-time, movement-coded auditory feedback effectively promotes motor skill learning in mice. This closed-loop system, leveraging advanced machine learning and real-time tracking, offers new avenues for exploring motor control mechanisms and developing therapeutic strategies for motor disorders through augmented sensory feedback.