Fig. 5From: Automated freezing of gait assessment with marker-based motion capture and multi-stage spatial-temporal graph convolutional neural networksAssessing the performance of the MS-GCN (6 stages) for automated FOG assessment. More specifically, the performance to measure the percentage time-frozen (%TF) (left) and the number of FOG episodes (#FOG) (right) during a standardized protocol. The ideal regression line with a slope of one and an intercept of zero is visualized in red. All results were derived from the test set, i.e., subjects that the model had never seen. Observe the overestimation of %TF and #FOG for S2Back to article page