Biomedical Engineering Reference
In-Depth Information
9.6.2 Box-lifting determinants
Based on the literature, six determinants—torso flexion [1], shoulder flexion [2],
elbow flexion [2,3], hip flexion [2,3,4], knee flexion [2,3,4], and ankle flexion
[2,3,4])—are identified in this work to determine if the PD model's (Santos s 's)
box-lifting motion correlates to the real-world observations.
9.6.3 Participants
Twelve subjects participated in the subjective evaluation of the predicted motion,
and another 12 subjects participated in the testing of a whole-body motion task.
The first 12 subjects were asked to watch movies of two avatars, one driven by
simulation data and one driven by experimental data, and then to evaluate the
avatars' motion normality. The second 12 subjects (8 healthy males and 4 healthy
females) were involved in the experimentation of the materials-lifting task. The
subjects had no history of musculoskeletal problems and were reasonably fit.
Their participation was voluntary, and a written informed consent, as approved by
the University of Iowa Institutional Review Board, was obtained prior to testing.
The height of
the subject population for
the materials-lifting task was
175.6
11.5 cm, the weight was 73.2
10.7 kg, and the age was 21.8
4.3 yrs.
6
6
6
9.6.4 Results
9.6.4.1 Qualitative comparison
Twelve subjects participated in the first benchmark test to observe and evaluate
two videos of animated avatars. One avatar was animated using experimental
data, and the other was animated using predicted data. The participants were
asked to report their scores based on what they consider a natural human lifting
motion. Figure 9.15 shows a comparison of the rating scores for both avatars. The
coefficient of correlation (r) between the experimental and simulation rate was
0.692729.
For the second benchmark, Figure 9.16A shows that the predicted
determinant-curve shapes, except for the ankle flexion, for lifting during the lift-
ing cycle have general shapes that closely followed the 12 individual subjects.
Figure 9.16B demonstrates the coupling strength between one of the predicted
determinants—shoulder flexion, for example—and other lifting determinants. As
shown in the figure, the predicted shoulder flexion determinant followed to some
extent the trend of coupling between the determinants of the natural subjects.
Again, the predicted ankle flexion showed weak correlation with the experiments.
Figure 9.17A shows a comparison between the velocity of the lifting determi-
nants of the 12 subjects and the model. As shown in the figure, the shape of the
predicted model velocity is similar to that of the subjects for the torso and hip.
However, there are fewer local fluctuations in the predicted lifting determinants.
The shape of the rest of the determinants was somehow different. The calculated
Search WWH ::




Custom Search