Artificial neural networks for analyzing inter-limb coordination: the golf chip shot

Area of Science:

• Motor control
• Biomechanics
• Computational neuroscience

Background:

• Motor control research often uses physical science theories like coordination dynamics.
• Previous studies focused on low-dimensional inter-limb coordination.
• High-dimensional movement coordination remains less understood.

Purpose of the Study:

• To investigate high-dimensional inter-limb coordination during golf chip shots.
• To introduce a novel visualization technique for coordination dynamics.
• To analyze changes in coordination patterns across different movement distances.

Main Methods:

• Utilized Artificial Neural Networks (ANN), specifically Self-Organizing Maps (SOM).
• Analyzed high-dimensional inter-limb coordination in golf chip shots towards six target distances.
• Employed SOM trajectories as collective variables for coordination stability analysis.

Main Results:

• SOM trajectories demonstrated shifts in coordination between short and long chip shots.
• Attractor diagrams revealed non-linear phase transitions in three out of four participants.
• The study successfully visualized high-dimensional coordination patterns.

Conclusions:

• The findings highlight distinct coordination strategies for varying golf chip shot distances.
• The proposed visualization method is effective for analyzing complex motor behaviors.
• This approach offers a valuable tool for coordination dynamics research on high-dimensional movements.