Table 1.
Summary of participant characteristics.
Figure 1.
A, illustration of recorded movements. Red lines indicate the direction of motion for joint angles included in the kinematic analysis. B, tracked points used to calculate joint angles. C, Average joint angles of the left arm calculated from the data captured by the two motion capture systems during 10 repetitions of the movement 8) shown in A.
Figure 2.
Results of principal component analysis.
Cumulative explained variance and the number of principal components are shown for each movement type across participants (top plot) and for each participant across movement types (bottom plot). Grey dotted lines show results of decomposition of movement of the non-paretic arm, while black solid lines shows results of decomposition of movement of the paretic arm. The principal components were derived from mean data and used to reconstruct data from individual movements.
Figure 3.
The comparison between quantitative scores from standard and low-cost motion capture and qualitative scores.
A, Dots show mean scores for each movement and each subject; thick line shows a regression fit. B, Symbols show mean scores for each subject; error bars show s.d. across 10 movements; thick line shows a regression fit.
Figure 4.
A, Error in predicting each subject's qualitative score from regressions fitted to the rest of the participants. Mean errors are expressed as % of the correct score; error bars show s.d. across 10 movements. B, Symbols show the same data as in Fig. 3B; lines show regressions for datasets with one subject's data point removed. C, Histogram of intraclass correlation coefficients for relationships between individual human raters and the mean qualitative score. D, Colored lines show reducing errors as more raters score movements of the same participants per movement type, limb, and participant.
Figure 5.
Relationships between qualitative and quantitative scores for each movement type.
Dots show mean scores across participants, thick lines show linear regressions with their equations and fit statistics above each plot.
Table 2.
Angular errors of low-cost motion capture relative to the standard system.