Presenter Type

UNO Graduate Student (Doctoral)

Major/Field of Study

Biomechanics

Other

Biomechanics

Author ORCID Identifier

https://orcid.org/0000-0003-3255-7252

Advisor Information

Assistant Professor

Location

MBSC Ballroom Poster # 1101 - G (Doctoral)

Presentation Type

Poster

Start Date

24-3-2023 9:00 AM

End Date

24-3-2023 10:15 AM

Abstract

Human movement is inherently variable by nature. One of the most common analytical tools for assessing movement variability is the largest Lyapunov exponent (LyE) which quantifies the rate of trajectory divergence or convergence in an n-dimensional state space. One popular method for assessing LyE is the Wolf algorithm. Many studies have investigated how Wolf’s calculation of the LyE changes due to sampling frequency, filtering, data normalization, and stride normalization. However, a surprisingly understudied parameter needed for LyE computation is evolution time. The purpose of this study is to investigate how the LyE changes as a function of evolution time in both simulated data and experimental data. The data from 36 healthy subjects were extracted from an ongoing study, investigating whether individuals possess a unique, self-identifying, gait. The subjects performed nine four-minute overground walking trials at their self-selected walking speed. Segment pitch angles from the left and right thigh, shank, and foot were extracted from the first 2-minutes of each trial to calculate LyEs. Simulated data consisted of values from a reconstructed Lorenz attractor. Evolution time was calculated by multiplying a fixed constant by the sampling rate and applying the ceiling function (ceil()) in MATLAB to round up to the nearest whole integer. Multi-level and linear models were used to assess whether the inclusion of fixed effects of evolution time improved prediction of LyE over and above an intercept only model for experimental and simulated data, respectively. Increasing evolution time in Wolf’s algorithm substantially negatively biases LyE values for simulated and experimental data. Overall, careful consideration should be taken when choosing the evolution time. An evolution time of 1.5 seconds produced the closest values to the expected simulated value. Therefore, future research should consider using a similar time for experimental data.

Scheduling

9:15-10:30 a.m., 10:45 a.m.-Noon, 1-2:15 p.m., 2:30 -3:45 p.m.

COinS
 
Mar 24th, 9:00 AM Mar 24th, 10:15 AM

TIME EVOLUTION IS A SOURCE OF BIAS IN THE WOLF ALGORITHM FOR LARGEST LYAPUNOV EXPONENTS

MBSC Ballroom Poster # 1101 - G (Doctoral)

Human movement is inherently variable by nature. One of the most common analytical tools for assessing movement variability is the largest Lyapunov exponent (LyE) which quantifies the rate of trajectory divergence or convergence in an n-dimensional state space. One popular method for assessing LyE is the Wolf algorithm. Many studies have investigated how Wolf’s calculation of the LyE changes due to sampling frequency, filtering, data normalization, and stride normalization. However, a surprisingly understudied parameter needed for LyE computation is evolution time. The purpose of this study is to investigate how the LyE changes as a function of evolution time in both simulated data and experimental data. The data from 36 healthy subjects were extracted from an ongoing study, investigating whether individuals possess a unique, self-identifying, gait. The subjects performed nine four-minute overground walking trials at their self-selected walking speed. Segment pitch angles from the left and right thigh, shank, and foot were extracted from the first 2-minutes of each trial to calculate LyEs. Simulated data consisted of values from a reconstructed Lorenz attractor. Evolution time was calculated by multiplying a fixed constant by the sampling rate and applying the ceiling function (ceil()) in MATLAB to round up to the nearest whole integer. Multi-level and linear models were used to assess whether the inclusion of fixed effects of evolution time improved prediction of LyE over and above an intercept only model for experimental and simulated data, respectively. Increasing evolution time in Wolf’s algorithm substantially negatively biases LyE values for simulated and experimental data. Overall, careful consideration should be taken when choosing the evolution time. An evolution time of 1.5 seconds produced the closest values to the expected simulated value. Therefore, future research should consider using a similar time for experimental data.