Stance and Swing Detection Based on the Angular Velocity of Lower Limb Segments During Walking

Abstract

Lower limb exoskeletons require the correct support magnitude and timing to achieve user assistance. This study evaluated whether the sign of the angular velocity of lower limb segments can be used to determine the timing of the stance and the swing phase during walking. We assumed that stance phase is characterized by a positive, swing phase by a negative angular velocity. Thus, the transitions can be used to also identify heel-strike and toe-off. Thirteen subjects without gait impairments walked on a treadmill at speeds between 0.5 and 2.1 m/s on level ground and inclinations between -10 and +10°. Kinematic and kinetic data was measured simultaneously from an optical motion capture system, force plates, and five inertial measurement units (IMUs). These recordings were used to compute the angular velocities of four lower limb segments: two biological (thigh, shank) and two virtual that were geometrical projections of the biological segments (virtual leg, virtual extended leg). We analyzed the reliability (two sign changes of the angular velocity per stride) and the accuracy (offset in timing between sign change and ground reaction force based timing) of the virtual and biological segments for detecting the gait phases stance and swing. The motion capture data revealed that virtual limb segments seem superior to the biological limb segments in the reliability of stance and swing detection. However, increased signal noise when using the IMUs required additional rule sets for reliable stance and swing detection. With IMUs, the biological shank segment had the least variability in accuracy. The IMU-based heel-strike events of the shank and both virtual segment were slightly early (3.3-4.8% of the gait cycle) compared to the ground reaction force-based timing. Toe-off event timing showed more variability (9.0% too early to 7.3% too late) between the segments and changed with walking speed. The results show that the detection of the heel-strike, and thus stance phase, based on IMU angular velocity is possible for different segments when additional rule sets are included. Further work is required to improve the timing accuracy for the toe-off detection (swing).

Keywords: control; detection; exoskeleton; gait phase; inertial measurement unit; stance; swing; walking.

Figure 1

Biological and virtual lower limb segments evaluated for the potential on stance and swing detection. The evaluated biological segments are the thigh (orange) and the shank (red). The evaluated virtual segments (based on Villarreal and Gregg, 2014) are the leg (blue) and the extended leg (green). While the virtual leg is a combination of thigh and shank, the virtual extended leg is a combination of the virtual leg and the thigh. Inertial measurement units (IMU) were place at the thigh and the shank to determine the segment angles and the segment angular velocities.

Figure 2

Lower limb segment velocity based concept for stance and swing detection. The shown angular velocity profile represents a conceived example case with similar timing of the angular velocity zero crossings from negative to positive at heel-strike, and the similar timing of the angular velocity zero crossings from positive to negative at toe-off. Heel-strike and toe-off can be determined by the existence vertical ground reaction force.

Figure 3

Measurement setup including an instrumented split-belt treadmill and a motion capture system (left). Inertial measurement unit (IMU) setup for the thigh and the shank (right). The sensor units were placed in textile straps. The length of the straps was adapted to differences in body composition using velcro. Reflective markers on the sensors were not used for this evaluation.

Figure 4

Examples for the applied rule sets. (A) The Event timer rule used a defined time interval after each event detection (heel-strike blue, toe-off green) to avoid additional hell-strike (yellow) or toe-off (red) detections. The example used an interval of 200 ms for the virtual leg during walking declines (1.7 m/s). (B) The Heel-strike rule ensured that after heel-strike only toe-off and after toe-off only heel-strike can be detected. In this example the Opposite leg rule is required in addition to also avoid the toe-off detection. (C) The Opposite leg rule required the heel-strike of the opposite limb (gray solid) before a toe-off can be detected. The example data for (B,C) is from the shank segment during walking inclines (0.9 m/s). (D) The Segment angle rule (was not applied for the analysis) required a change of 15° in the segment angle (set to zero when angular velocity larger than −50 °/s) to allow toe-off (green) detection. In combination with the Heel-strike rule it can avoid toe-off detection while standing. The example data is from the shank segment during gait initiation of walking inclines. Heel-strike is set as initial condition. The timing of the stance and the swing, indicated by the bottom bars for (A–D) is based on the rule sets.

Figure 5

Angular velocities from inertial measurement units (solid) and motion capture (dashed) for the thigh (orange), shank (red), leg (blue), and extended leg segment (green) in level walking, inclined walking (10°) and declined walking (10°). Darker colors indicate higher speeds (0.9, 1.3, and 1.7 m/s for slopes, 0.5, 1.3, and 2.1 m/s for level walking). Circles indicate the related toe-off determined by the vertical ground reaction force. The heel-strike and the following heel-strike of the same limb occur at 0 and 100% of the stride time. Gray areas indicate the swing phase based on the angular velocity of the IMU at 1.3 m/s.

Figure 6

Number of zero crossings of the angular velocity per gait cycle for the thigh (orange), shank (red), leg (blue), and extended leg segment (green) in level walking, inclined walking (10°) and declined walking (10°). The angular velocity was determined by motion capture (left) and by inertial measurement units (right). Darker colors represent greater speeds (0.9, 1.3, and 1.7 m/s for slopes, 0.5, 1.3, and 2.1 m/s for level walking). Error bars represent one standard deviation.

Figure 7

Single subject angular velocities for one stride from the inertial measurement units (solid) and the motion capture (dashed) for the thigh (orange, descent −10° at 1.7 m/s), shank (red, ascent 10° at 0.9 m/s), leg (blue, descent −10° at 1.7 m/s), and extended leg (green, descent −10° at 1.7 m/s) segment. Circles indicate phases where the IMU data can have unwanted sign changes in the angular velocity due to oscillations or by the nature of the segment movement. Representative strides were selected from one subject and condition with the greatest number of zero crossings shown in Figure 6.

Figure 8

Time difference of the mean zero crossings of the angular velocity based on the optical motion capture data to the heel-strike (left) and the toe-off (right) identified by ground reaction forces (GRF). Evaluated segment angular velocities from the thigh (orange), shank (red), leg (blue), and extended leg (green). Distances are evaluated for three different speeds of level walking, walking inclines, and walking declines. Darker colors indicate greater speeds (0.9, 1.3, and 1.7 m/s for slopes, and 0.5, 1.3, and 2.1 m/s for level walking). The standard deviation is indicated by the vertical line. The specific detection rule sets, that were designed for the IMU data, were also applied to the Motion capture data to only detect one transition for each event. Positive and negative values indicate zero crossings after and before, respectively, the GRF based event detection.

Figure 9

Time difference of the mean zero crossings of the angular velocity based on the inertial measurement unit to the heel-strike (left) and the toe-off (right) identified by ground reaction forces (GRF). Evaluated segment angular velocities from the thigh (orange), shank (red), leg (blue), and extended leg (green). Distances are evaluated for three different speeds of level walking, walking inclines, and walking declines. Darker colors indicate greater speeds (0.9, 1.3, and 1.7 m/s for slopes, and 0.5, 1.3, and 2.1 m/s for level walking). The standard deviation is indicated by the vertical line. As multiple zero crossings occur for all segments, specific detection rule sets were designed to only detect one transition for each event. Positive and negative values indicate zero crossings after and before, respectively, the GRF based event detection.