Affine differential geometry analysis of human arm movements

Area of Science:

• Neuroscience
• Mathematics
• Biomechanics

Background:

• Human interaction with the environment relies on sensory information and motor actions.
• Observed regularities in human movement, such as the two-thirds power law and local isochrony, suggest non-Euclidean geometry in motor control.
• Current understanding often defaults to Euclidean geometry for motor spaces, potentially overlooking crucial underlying structures.

Purpose of the Study:

• To develop a mathematical framework for analyzing human hand trajectories based on empirical observations.
• To identify potential elementary building blocks (motor primitives) of complex movements.
• To investigate the geometric structure underlying human motion perception and production.

Main Methods:

• Utilized differential geometry, Lie group theory, and Cartan's moving frame method to analyze hand trajectories.
• Developed an equi-affine geometric framework describing movement characteristics.
• Calculated equi-affine arc-length and curvature for experimentally recorded drawing movements.

Main Results:

• Demonstrated that human hand trajectories are best described by equi-affine geometry, not Euclidean.
• Equi-affine velocity is piecewise constant, and movement time is proportional to equi-affine arc-length.
• Identified conic sections, particularly parabolas, as potential motor primitives and equi-affine geodesics.

Conclusions:

• The equi-affine framework provides a natural geometric description of continuous repetitive hand trajectories.
• This geometric approach is compatible with neural coding models for motor commands.
• The equi-affine metric may play a significant role in the internal representations of motion perception and production.