Skip to main content

Table 5 Results of the stride segmentation performance, for the free-living dataset. All values are given as mean ± std across the 28 subjects, over all strides per day, per patient

From: Hidden Markov Model based stride segmentation on unsupervised free-living gait data in Parkinson’s disease patients

Method Precision [%] Recall [%] F1-Score [%]
DTW \(87.2 \pm 8.4\) \(83.3 \pm 10.1\) \(85.1 \pm 9.0\)
\(\hbox {HMM}_{\mathrm{lab}}\) \(89.4 \pm 5.6\) \(95.2 \pm 3.0\) \(92.2 \pm 4.2\)
\(\hbox {HMM}_{\mathrm{free}\hbox {-}\mathrm{living}}\) \(89.4 \pm 5.7\) \(95.6 \pm 2.8\) \(92.4 \pm 4.1\)