Open Collections

Rong, Shenyi

University of British Columbia

Inertial measurement units (IMUs) are a prevalent component in wearable sensor systems to study motion in unconstrained environments by measuring both linear acceleration and angular velocity at a point of interest on the human body. However, wearable sensors are often worn on soft tissue that is excited during impulsive impacts, producing dynamic soft tissue artifact measurement errors. This thesis quantifies the error caused by dynamic soft tissue artifacts to IMU sensors with different sensor mass and placement when mild impulsive impacts are delivered to the lower extremity using empirical signal analysis and model-based predictions. Data were collected from three IMU sensors attached at the anterior and posterior center of the dominant shank, and at the back of the ankle of 10 participants (5 females, 5 males). We varied the mass of the two shank IMUs from 5 grams to 37 grams in 8-gram increments. Impacts were generated by dropping the shank freely onto a foam pad placed on a force plate from a height of 21.6 cm. Ground truth vertical linear acceleration measurements from the ankle IMU were first cross-validated against the force plate. We then determined both posterior and anterior IMUs overestimated the ground truth peak vertical linear acceleration measurement by 117.9% and 55.1%, respectively. The posterior IMU had greater overestimations, suggesting more severe dynamic soft tissue artifacts, while errors generally decreased with increasing mass, suggesting a mitigating effect from adding sensor mass. We next assessed post-impact oscillations, which we determined lasted 0.32±0.07 seconds with a frequency of 9.79±2.68 Hz on the posterior and 0.15±0.05 seconds with a frequency of 18.33±12.24 Hz on the anterior. As sensor weight increased, the oscillation duration increased linearly while the frequency decreased for both anterior and posterior IMUs. This suggests that increasing sensor mass amplifies post-impact dynamic soft tissue artifact errors. These results and dependencies were confirmed by modeling dynamic soft tissue artifacts as double spring mass damper systems. In conclusion, the dynamic soft tissue artifacts directly relate to the accuracy of kinematic measurement in IMU sensors with a dependence on sensor mass and placement.

Text

2022-10-19

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/82903

Biomedical Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/