Automated detection of missteps during community ambulation in patients with Parkinson’s disease: a new approach for quantifying fall risk in the community setting

Table 3 Features extracted from the 3 days recordings

Feature	Axis	Suspected misstep		Not suspected misstep
Feature	Axis	Mean	Std	Mean	Std
Peak difference	Vertical [g]	0.59	0.23	0.57	0.23
	Medio-lateral [g]	0.60	0.20	0.60	0.20
	Anterior- posterior [g]	0.56	0.20	0.55	0.23
	Yaw [deg/sec]	0.68	0.18	0.66	0.19
	Pitch [deg/sec]	0.64	0.21	0.63	0.20
	Roll [deg/sec]	0.65	0.20	0.64	0.20
Frequencies above threshold	FFT- Vertical	6.42	3.17	1.50	0.93
Entropy	FFT- Vertical	2.46	0.34	1.39	0.45
Number of steps	Vertical	10.58	1.47	8.10	3.08
Maximum amplitude [g]	Vertical	0.98	0.56	0.54	0.43
Number of steps	Anterior- posterior	9.94	0.99	6.84	2.81
Maximum amplitude [g]	Anterior- posterior	1.34	0.5	0.43	0.35
Number of steps	Yaw	4.58	3.3	4.5	3.58
Maximum amplitude [deg/sec]	Yaw	1.99	0.54	1.57	0.62

The division between sMS or not for each feature is according to the threshold of the feature.

ISSN: 1743-0003