Exact box-counting and temporal sampling algorithms for fractal dimension estimation with applications to animal...

Area of Science:

• Quantitative Biology
• Biophysics
• Computational Neuroscience

Background:

• Fractal dimension (FD) quantifies complexity in natural systems.
• Assessing animal movement complexity is crucial for understanding behavior and disease.
• Existing methods for FD estimation may lack accuracy with high-resolution movement data.

Purpose of the Study:

• To develop novel algorithms for accurate fractal dimension estimation of animal movement.
• To introduce methods for handling spatial and temporal sampling in movement data.
• To investigate the impact of the Dysbindin gene on movement complexity in Drosophila larvae.

Main Methods:

• Developed an oversampling technique using linear interpolation for movement paths.
• Introduced an exact box-counting algorithm for piecewise linear paths.
• Proposed a temporal sampling method for calculating FD in the temporal domain.
• Employed dual total least squares for optimal scale selection in FD comparisons.

Main Results:

• The novel algorithms provide accurate FD estimations for animal movement.
• The study identified significant increases in FD for Drosophila larvae with Dysbindin mutations.
• FD was shown to be a sensitive metric for detecting motor function alterations.

Conclusions:

• Fractal dimension is a robust metric for quantifying animal movement complexity.
• Dysbindin gene mutations are associated with increased movement complexity, suggesting motor impairments.
• The developed algorithms enhance the application of FD analysis in behavioral studies.