Enhanced spatio-temporal alignment of plantar pressure image sequences using B-splines

Area of Science:

• Biomechanics
• Medical Imaging
• Computer Vision

Background:

• Accurate alignment of plantar pressure image sequences is crucial for analyzing gait dynamics.
• Existing methods for temporal and spatial alignment have limitations in precision and consistency.
• Plantar pressure data provides valuable insights into foot function during locomotion.

Purpose of the Study:

• To present an enhanced methodology for simultaneous temporal and spatial alignment of plantar pressure image sequences.
• To evaluate the accuracy and effectiveness of cubic B-splines for temporal alignment.
• To assess the impact of temporal alignment on the consistency of dynamic center of pressure (COP) displacement.

Main Methods:

• Employed B-splines for temporal modeling and geometric transformations for spatial alignment.
• Tested the methodology on 156 real plantar pressure image sequences from barefoot walking.
• Assessed alignment accuracy using synthetic deformations and evaluated COP displacement consistency via intraclass correlation coefficients (ICC).

Main Results:

• Cubic B-splines achieved outstanding accuracy in aligning synthetically deformed image sequences, significantly outperforming previous methods (p < 0.001).
• Superior alignment results were observed for real image sequences with unknown transformations compared to prior methodologies (p < 0.001).
• Temporal alignment generally increased the intraclass correlation coefficients (ICCs) for medio-lateral COP displacement, indicating enhanced consistency.

Conclusions:

• Cubic B-splines offer a highly effective solution for the temporal alignment of plantar pressure image sequences.
• The enhanced temporal alignment methodology improves the reliability and consistency of dynamic COP displacement analysis.
• This improved alignment has significant implications for accurate gait analysis and clinical biomechanics research.