0
Research Papers

An Automated Image-Based Method of 3D Subject-Specific Body Segment Parameter Estimation for Kinetic Analyses of Rapid Movements

[+] Author and Article Information
Alison L. Sheets

Department of Mechanical Engineering, Biomotion Laboratory, Stanford University, Stanford, CA 94305alsheets@stanford.edu

Stefano Corazza

Department of Mechanical Engineering, Biomotion Laboratory, Stanford University, Stanford, CA 94305

Thomas P. Andriacchi1

Department of Mechanical Engineering, Biomotion Laboratory, Stanford University, Stanford, CA 94305; Department of Orthopedics, Stanford University, Stanford, CA 94305; Palo Alto Veterans Affairs, Bone and Joint Center, Palo Alto, CA 94304

1

Corresponding author.

J Biomech Eng 132(1), 011004 (Dec 09, 2009) (10 pages) doi:10.1115/1.4000155 History: Received January 26, 2009; Revised July 02, 2009; Posted September 03, 2009; Published December 09, 2009; Online December 09, 2009

Accurate subject-specific body segment parameters (BSPs) are necessary to perform kinetic analyses of human movements with large accelerations, or no external contact forces or moments. A new automated topographical image-based method of estimating segment mass, center of mass (CM) position, and moments of inertia is presented. Body geometry and volume were measured using a laser scanner, then an automated pose and shape registration algorithm segmented the scanned body surface, and identified joint center (JC) positions. Assuming the constant segment densities of Dempster, thigh and shank masses, CM locations, and moments of inertia were estimated for four male subjects with body mass indexes (BMIs) of 19.7–38.2. The subject-specific BSP were compared with those determined using Dempster and Clauser regression equations. The influence of BSP and BMI differences on knee and hip net forces and moments during a running swing phase were quantified for the subjects with the smallest and largest BMIs. Subject-specific BSP for 15 body segments were quickly calculated using the image-based method, and total subject masses were overestimated by 1.7–2.9%.When compared with the Dempster and Clauser methods, image-based and regression estimated thigh BSP varied more than the shank parameters. Thigh masses and hip JC to thigh CM distances were consistently larger, and each transverse moment of inertia was smaller using the image-based method. Because the shank had larger linear and angular accelerations than the thigh during the running swing phase, shank BSP differences had a larger effect on calculated intersegmental forces and moments at the knee joint than thigh BSP differences did at the hip. It was the net knee kinetic differences caused by the shank BSP differences that were the largest contributors to the hip variations. Finally, BSP differences produced larger kinetic differences for the subject with larger segment masses, suggesting that parameter accuracy is more important for studies focused on overweight populations. The new image-based BSP estimation method described in this paper addressed the limitations of currently used geometric and regression methods by using exact limb geometry to determine subject-specific parameters. BSP differences have the largest effect on kinetic analyses of motions with large limb accelerations, for joints farther along the kinematic chain from the known forces and moments, and for subjects with larger limb masses or BMIs.

FIGURES IN THIS ARTICLE
<>
Copyright © 2010 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Figure 1

An illustration of the steps used for image-based BSP calculation: (a) laser scan subject, (b) segment the laser-scanned surface and identify joint center locations, and (c) fill segment ends and assume uniform density to calculate each limb’s BSP

Grahic Jump Location
Figure 2

Steps in the iterative pose shape registration algorithm. A segmented, rigid body generic model and a laser-scanned surface are input. The model’s relative segment orientations are adjusted and the outer surface is morphed to match the laser scan. A segmented subject-specific model is output (27).

Grahic Jump Location
Figure 3

Shank and thigh free body diagrams

Grahic Jump Location
Figure 4

Left thigh (a) and shank (b) mass normalized by total body mass for each subject calculated using each method

Grahic Jump Location
Figure 5

Frontal (top row) and sagittal (bottom row) plane view of the left thigh and left shank CM locations calculated using each method for S1 and S4, which are the subjects with the smallest and largest BMI, respectively

Grahic Jump Location
Figure 6

Left thigh (a) and shank (b) medial-lateral moments of inertia about the CM normalized by total body mass multiplied by height squared (kg m2) for each subject calculated using each method

Grahic Jump Location
Figure 7

S1 (BMI=19.7, top row) and S4 (BMI=38.2, bottom row) net knee (left column) and hip (right column) distal-proximal forces in the thigh reference frame during running swing phase

Grahic Jump Location
Figure 8

S1 (BMI=19.7, left) and S4 (BMI=38.2, right) net knee flexion/extension moments during running swing phase

Grahic Jump Location
Figure 9

S1 (BMI=19.7, top row) and S4 (BMI=38.2, bottom row) contributions to net knee and hip anterior-posterior (A-P) and distal-proximal (D-P) forces in the thigh reference frame at the instants when peak M-L knee and hip moments occur during running swing phase (Eqs. 1,3)

Grahic Jump Location
Figure 10

S1 (BMI=19.7, top row) and S4 (BMI=38.2, bottom row) force and moment contributions to peak net medial-lateral knee and hip moments during running swing phase (Eqs. 2,4)

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In