Feature Extraction

Several features can be extracted from the segmented data. These features can be used to analyze the gait pattern of the subject. The features are calculated for each segment of the trial. The features are calculated using the gaitalytics.api.calculate_features() function.

from gaitalytics import api

config = api.load_config("./config.yaml")
trial = api.load_c3d_trial("./example_with_events.c3d", config)
segmented_trial = api.segment_trial(trial)
features = api.calculate_features(segmented_trial, config)

The function returns a DataArray object with the coordinates feature, cycles, context

Features

Following features can be extracted from the segmented data:

Method	Description	Unit
TimeSeriesFeatures	min	deg, Nmm, N, V*
	max	deg, Nmm, N, V*
	mean	deg, Nmm, N, V*
	sd	deg, Nmm, N, V*
	median	deg, Nmm, N, V*
	amplitude	deg, Nmm, N, V*
PhaseTimeSeriesFeatures	stand_min	deg, Nmm, N, V*
	stand_max	deg, Nmm, N, V*
	stand_mean	deg, Nmm, N, V*
	stand_sd	deg, Nmm, N, V*
	stand_median	deg, Nmm, N, V*
	stand_amplitude	deg, Nmm, N, V*
	swing_max	deg, Nmm, N, V*
	swing_mean	deg, Nmm, N, V*
	swing_sd	deg, Nmm, N, V*
	swing_median	deg, Nmm, N, V*
	swing_amplitude	deg, Nmm, N, V*
SpatialFeatures	step_length	cm, mm✝
	stride_length[1]	cm, mm✝
	step_length[1]	cm, mm✝
	step_width[1]	cm, mm✝
	minimal_toe_clearance[2]	cm, mm✝
	AP_margin_of_stability[3]	cm, mm✝☨
	ML_margin_of_stability[3]	cm, mm✝☨
TemporalFeatures	cycle_duration	sec
	double_support[4]	%
	single_support[4]	%
	stance_duration_prec	%
	swing_duration_prec	%
	opposite_foot_off_prec	%
	opposite_foot_contact_prec	%
	stride_duration	sec
	stance_duration	sec
	cadence[1]	steps/min

Foot note

* Depending on the unit in the c3d file.
✝ Depending on the unit of the configured entity in the c3d file.
☨ Feature only calculated if a center of mass marker is present in the c3d file and configured in the config file.

References