Title: Determining Validity of Using Inertial Measurement Units to Calculate Ground Reaction Force
Name: Kelsey Wagner
Primary Presenters: Amy Jensen, SPT
College of Health Sciences
Clinical Doctorate
Faculty Sponsor(s): Wendy Huddleston, PT, PhD

Background: Anterior cruciate ligament (ACL) tears can end an athletic career and predispose individuals to osteoarthritis (Almonroeder et al., 2018). Ground reaction forces (GRFs) are used to determine the risk of an ACL tear (Almonroeder et al., 2018). Force plates are the gold standard for determining GRFs, limiting research to the laboratory. Inertial measurement units (IMUs) can be worn to track acceleration data in the field. Theoretically, acceleration data can be multiplied by the subject’s mass to find GRF using Newton’s second law of motion.
Purpose: Determine if IMUs validly measure GRFs.
Hypothesis: IMUs can produce valid values for GRFs in the field.
Subjects: Four researchers aged 23 to 25 (n=3 females, 1 male) were used for data collection.
Materials & Methods: Mass (kg) of each participant was collected. Participants wore well-fitted running shoes. An IMU was placed using an adhesive sticker at the L4-L5 spinal segment (Setuain et al., 2016) that was palpated by one researcher and confirmed by another. A 31 cm tall box was placed 15.24 cm behind the force plate (Earl et al., 2007 as cited in Almonroeder et al., 2018). Participants then performed a single leg step off the box onto the force plate with their feet 35 cm apart (Almonroeder et al., 2018). Each participant repeated the single leg drop three times for each leg. Data was collected simultaneously using the IMU and the force plate. The force plate provided data in Newtons (N). IMU data was converted using F=ma (Force, mass, peak acceleration in m/s 2) (Setuain et al., 2016). Data sets were averaged for each leg. IMU and force plate data were assessed for correlation. Following the calculation of the correlations, a permutation was run to further examine the strength of the correlation values.
Results: There was a moderate positive correlation between the IMU and force plate data in the vertical direction (r=0.5552) and the side-to-side direction (r=0.5559). A weak positive correlation in the forward-backward direction was found (r=0.3667). To further assess the data, a one tail t-test permutation was performed producing no significant differences in the vertical (401/2068 permutations had better correlation, r=0.364, p=0.194), side-to-side (380/2064 permutations had better correlation, r=0.374, p=0.184) and forward-backward (172/2052 permutations had a better correlation, r=0.553, p=0.084) directions, further increasing our confidence in the correlations calculated and the conclusions drawn from them.
Conclusion: Results from the current pilot study are inconclusive regarding the validity of using IMUs to measure GRFs. Further research should be done using a larger sample size or different IMU placement to better determine if IMU can be used to validly determine GRFs outside the laboratory setting.