Impact of Neural network-quantified musical groove on cyclists' joint coordination and muscle synergy: a repeated...

While music is known to influence exercise performance through auditory-motor coupling, the biomechanical mechanisms by which musical groove-characterized by rhythmic drive and movement-inducing qualities-modulates cycling coordination remain poorly understood. This study investigates how groove levels systematically alter lower extremity kinematics and neuromuscular control during high-torque cycling.

Twenty-four well-trained, right-handed cyclists completed high-torque cycling trials under three counterbalanced conditions: metronome (control), low-groove (LG), and high-groove (HG) music, with groove levels objectively classified by a validated deep learning model (R² = 0.85). Three-dimensional motion capture (200Hz) quantified hip-ankle and pelvis-torso coordination using vector coding techniques, while surface electromyography (EMG) of 12 lower limb muscles was analyzed via non-negative matrix factorization (NMF) to extract muscle synergy patterns. The NMF approach decomposes multi-muscle activation patterns into fundamental synergistic components, providing insight into neuromuscular control strategies.

Compared to LG and control conditions, HG music significantly: (1) increased hip-ankle in-phase coordination by 28.7% (HG:29.8% vs. LG:23.2%, p = 0.020), (2) enhanced pelvis-torso synchronization by 27.1% (HG:38.0% vs. LG:29.9%, p = 0.048), and (3) promoted greater muscle synergy complexity (median synergies: HG = 7 vs LG = 6, p = 0.039). Notably, the soleus (SOL) muscle-crucial for ankle stabilization-showed significantly higher activation weights in HG condition (0.11 ± 0.03 vs 0.04 ± 0.02, p = 0.030), suggesting improved distal control. The emergence of a unique erector spinae-gastrocnemius lateralis (ES-GL) synergy pattern (present in 54% of HG trials) indicates enhanced trunk-limb coupling under high-groove conditions.

High-groove music promotes more coordinated movement patterns during cycling through two key mechanisms: (1) optimized joint coordination, particularly in proximal-distal (hip-ankle) and axial (pelvis-torso) linkages, and (2) reorganization of neuromuscular control strategies evidenced by increased synergy complexity and selective activation of postural stabilizers (SOL). These findings provide biomechanical evidence supporting groove-based auditory-motor interventions, though direct performance benefits require verification through additional kinetic and metabolic measures. The successful application of deep learning for groove quantification establishes a framework for personalized music selection in sports and rehabilitation contexts.