Butterfly Plots Tracking a Person’s Center of Pressure

Researcher(s)

  • Sophia Burrowes, Biomedical Engineering, University of Delaware

Faculty Mentor(s)

  • Dr. Jill Higginson, Mechanical Engineering, Biomedical Engineering, University of Delaware

Abstract

Post-stroke patients have an asymmetric gait which means an increased chance of falling. Center of pressure (COP) or butterfly plots and calculated gait symmetry will provide quantitative data displaying rehabilitation progress and an efficient way to understand a patient’s gait characteristics. 

Participants walked on an adaptive split-belt treadmill (ATM) and an adaptive tied-belt treadmill ( Pariser 2022) which updated each belt speed independently based on the individual’s speed on that side. There were five randomized trials where participants were to maintain the same speed while one or both belts required various levels of propulsion. The conditions were TiedNormal, Tied Hard, SplitNormal, SplitHard, and SplitUneven. “Tied” mode where Belt 1 speed= Belt 2 speed. “Split” where each leg controlled a belt. “Normal” where propulsion was similar to overground walking. “Hard” where more propulsion was required to maintain  their original speed. “Uneven” where Belt 1=Normal and Belt 2= Hard. 

Ground reaction forces were collected on an instrumented Bertec treadmill to calculate COP which is a weighted average of all pressures in contact with the ground. The COP plot is a shape similar to a butterfly. Average gait symmetry was calculated as abs(a-b)/(a+b), where “a” is the length between the left maximum point to the right side’s minimum point and “b” is the length of the right maximum to the left minimum. Force Data from 14 participants was processed and analyzed with custom MATLAB code. We ran a one-way repeated measures ANOVA and follow-up Wilcoxon tests with Bonferroni correction using python. We found a significant effect of the trials on the symmetry. The results showed that there was a significant difference in gait symmetry between SplitUneven and TiedHard which supported our hypothesis. 

 This code is applicable for impaired individuals and could be useful in clinical settings as it creates a summary report about a person’s symmetry and the plot encapsulates key gait characteristics in one image.