Influence Of Sex On Marker To Marker-less Motion Capture Evaluations Of Gait Biomechanics

Marker-based (MB) motion capture is affected by pelvic shape and soft tissue distribution that differ between sexes. Marker-less (ML) motion capture may provide an alternative. PURPOSE: To compare the effect of sex and method on knee and hip kinematics during gait. METHODS: Twenty participants (10 females: Age = 20.3 ± 0.95 years; Body Mass Index = 21 ± 1.83 kg/m2; 10 males: Age = 21.7 ± 1.34 years; Body Mass Index = 23.37 ± 1.37 kg/m2) participated. Participants completed two over-ground gait sessions at a self-selected pace with MB and ML recording a full right limb gait cycle simultaneously. Hip and knee kinematics in the sagittal and frontal plane were extracted and time-normalized from 0-100% of the gait cycle. Statistical parametric mapping using 2 (sex) by 2 (method) ANOVA with repeated measures compared kinematic waveforms. RESULTS: Interaction effects were found for sagittal plane hip angle from 0-7% (p = 0.009) and 71-100% (p = 0.004) of gait (F = 9.013). ML hip flexion angle was lower for 100% of gait compared to MB (p > 0.001, Z = 2.69), but sex comparisons were influenced by method. Females had greater hip flexion than males in MB (p < 0.001, Z = 3.29) from 0-33% and 64-100% of gait but no difference between sexes were found in ML. There was a main effect of method for frontal plane hip angle from 57-79% of gait (p = 0.007, F = 9.728). ML had less hip abduction (p = 0.005, Z = 2.92) from 57-79% of gait compared to MB. A main effect of method was found for sagittal plane knee angle from 3-23% and 39-100% of gait (p < 0.001, F = 11.965) and frontal plane knee angle from 34-40% and 84-92% of the gait cycle (p = 0.033, p = 0.019, F = 11.261). Knee flexion from 1-25% and 38-100% (p < 0.001, Z = 3.11) and knee adduction from 27-41% and 83-93% of gait were less for ML compared with MB, respectively (p = 0.01, p = 0.003, Z = 3.12). A main effect of sex was found for frontal plane hip angle from 0-40% and 63-100% (p = 0.001, F = 9.013). Females had a larger hip adduction angle than males (p = 0.007, Z = 2.86) between 72-90% of gait. CONCLUSION: Motion capture method influences gait comparisons between sexes, particularly at the hip joint. Studies are needed to develop sex specific ML models that identify differences in hip kinematics between males and females. Lastly, consistent main effects were observed of method that indicate some offset between MB and ML methods for other outcomes. Canadian Foundation for Innovation John R. Evans Leader Fund (Project 42110; PI: Pamukoff) Natural Science and Engineering Research Council of Canada (RGPIN-2022-04804; PI: Pamukoff)