Sampling frequency influences sample entropy of kinematics during walking

Area of Science:

• Biomechanics
• Kinematics
• Complexity Science

Background:

• Sample entropy (SaEn) quantifies gait regularity using joint angle data.
• Previous studies have not fully explored the impact of sampling frequency and stride count on SaEn calculations for lower limb kinematics.

Purpose of the Study:

• To investigate how variations in sampling frequency and the number of strides affect Sample Entropy (SaEn) calculations for lower limb joint angles during walking.
• To determine the optimal parameters for reliable SaEn analysis in gait kinematics.

Main Methods:

• Eleven subjects underwent 10-minute treadmill walking sessions at their preferred speed.
• Sagittal plane hip, knee, and ankle joint angle data were collected.
• SaEn was computed using stride counts of 50, 100, 200, 300, and 400 at sampling frequencies of 60, 120, 240, and 480 Hz.

Main Results:

• Increased sampling frequency significantly decreased SaEn values for all three lower limb joints.
• The number of strides included had a negligible effect on hip joint SaEn and a limited effect on knee and ankle joint SaEn.
• SaEn is more influenced by the number of data points within each stride than the total number of strides analyzed.

Conclusions:

• Sampling frequency is a critical factor influencing SaEn in lower limb joint angle signals during walking.
• Researchers should standardize the number of data points per stride, rather than the number of strides, for consistent SaEn analysis.
• This finding is crucial for accurate and reproducible gait complexity assessments.