Wearable IMMU-Based Relative Position Estimation between Body Segments via Time-Varying Segment-to-Joint Vectors.

Area of Science:

• Biomechanics
• Human Motion Analysis
• Wearable Technology

Background:

• Estimating relative positions between body segments using IMMUs is crucial for unconstrained human motion capture.
• Current methods assume rigid body segments, using constant segment-to-joint center (S2J) vectors, which leads to inaccuracies due to soft tissue artifacts (STAs).

Purpose of the Study:

• To develop a novel method for determining time-varying S2J vectors to enhance the accuracy of IMMU-based relative position estimation.
• To compensate for soft tissue artifacts (STAs) that affect human body segment rigidity.

Main Methods:

• A regression method was used to derive a function that predicts S2J vectors based on physical quantities correlated with segment deformation.
• This function was then used to determine time-varying S2J vectors, reflecting real-time segment deformation.

Main Results:

• The proposed method achieved an average root mean squared error of 15.08 mm, significantly outperforming the conventional method using constant vectors (31.32 mm).
• Validation across three subjects and four tests demonstrated the superior accuracy of the time-varying S2J vector approach.

Conclusions:

• The proposed method effectively compensates for STAs, leading to more accurate relative position estimations between body segments.
• This technique, applied in wearable motion tracking systems, holds potential benefits for applications in rehabilitation and sports sciences.