Predictive regression modeling of body segment parameters using individual-based anthropometric measurements

Area of Science:

• Biomechanics
• Ergonomics
• Medical Imaging

Background:

• Body segment parameters (BSPs) are crucial inputs for biomechanical models predicting joint/muscle forces and injury risks.
• Existing BSP prediction methods have significant errors (up to 40-60%), limiting model accuracy.
• Accurate BSPs are needed for reliable prediction of musculoskeletal injury risks.

Purpose of the Study:

• Develop and validate statistical models for predicting BSPs in working adults.
• Utilize whole-body DXA scan data and anthropometric measurements for improved BSP prediction.
• Enhance the accuracy of segment parameter inputs for biomechanical and ergonomic modeling.

Main Methods:

• Developed statistical models using a training dataset of working adults.
• Incorporated whole-body dual-energy X-ray absorptiometry (DXA) scan data and anthropometric measurements.
• Validated models on an independent test dataset to assess prediction accuracy.

Main Results:

• New models predicted BSPs within 5% of in vivo DXA-based measurements on average.
• Significantly improved accuracy compared to previous methods (average errors up to 60%).
• Demonstrated high accuracy for predicting torso, thigh, shank, upper arm, and forearm segment parameters.

Conclusions:

• The developed statistical models provide highly accurate and representative BSPs for individuals.
• These models enhance the reliability of biomechanical and ergonomic modeling outputs.
• Improved BSP prediction facilitates more accurate assessment of joint/muscle forces and injury risks.