Figure 5
From: Machine learning classifies predictive kinematic features in a mouse model of neurodegeneration
Feature scores for each anatomical location in different datasets. Four-month-only and 13mon-only datasets yielded different cumulative FI scores (%) for different anatomical locations, suggesting that kinematic synergy (or the lack thereof) during locomotion was correlated with age and pathological progression. Each bubble shown represents an anatomical location analyzed in this study. The size of the bubble is proportionate to its cumulative FI score calculated from bilateral or unilateral datasets of (a) the 4mon-only group, and (b) the 13mon-only group. The cumulative FI scores (%) were calculated by summing all the FI scores related to each anatomical location. All 82 features extracted from all body parts were included. As the mice aged, more joints become important for differentiating the two genotypes instead of just the hip, shoulder, and elbow (c). The left–right comparison was also performed for 4mon (d) and 13mon (e) respectively. Panels generated through MATLAB 9.2 and Adobe Illustrator (2019).
