Smart Shoe-Assisted Evaluation of Using a Single Trunk/Pocket-Worn Accelerometer to Detect Gait Phases

Abstract

Wearable sensors may enable the continuous monitoring of gait out of the clinic without requiring supervised tests and costly equipment. This paper investigates the use of a single wearable accelerometer to detect foot contact times and estimate temporal gait parameters (stride time, swing and stance duration). The experiments considered two possible body positions for the accelerometer: over the lower trunk and inside a trouser pocket. The latter approach could be implemented using a common smartphone. Notably, during the experiments, the ground truth was obtained by using a pair of sensorized shoes. Unlike ambient sensors and camera-based systems, sensorized shoes enable the evaluation of body-worn sensors even during longer walks. Experiments showed that both trunk and pocket positions achieved promising results in estimating gait parameters, with a mean absolute error below 50 ms.

Keywords: accelerometer; foot contact detection; gait analysis; gait phase detection; pocket-worn; smart shoe; wearable sensor.

Figure 1

Gait cycle and its phases. A gait cycle is defined as the interval between consecutive heel-strike (HS) events of the same leg. The duration of a gait cycle is also known as stride time. The toe-off (TO) event defines the two phases of a leg during a gait cycle: stance (foot on the ground), and swing (foot swinging after toe-off and before heel-strike).

Figure 2

Shimmer3 devices (left) and FootMoov 2.0 shoes (right).

Figure 3

Device placement during experiments (Shimmer3 at trunk and pocket positions; FootMoov 2.0 shoes) and reference anatomical directions. In particular, the anterior-posterior (AP) direction is aligned with the direction of motion during gait.

Figure 4

Detection of gait phases with the sensorized shoe. The thresholds TH_TO₁ and TH_TO₂ are used to detect a transition from stance to swing (TO event). A single threshold TH_HS is used to detect a transition from swing to stance (HS event). Thresholds were determined experimentally.

Figure 5

Example of force sensor signals on a FootMoov shoe during gait. Shaded areas highlight stance intervals.

Figure 6

Detection of gait parameters at the trunk position. In (a): shaded areas highlight the intervals identified by the walking detection algorithm based on acceleration magnitude analysis; squares indicate HS events, circles indicate TO events; red color is used for the events of one leg, black dashed for the other one. In (b) the estimated parameters are shown for the leg making the first step in this example: shaded areas highlight stance periods.

Figure 7

Detection of gait parameters at the pocket position. In (a): shaded areas highlight the intervals identified by the walking detection algorithm based on acceleration magnitude analysis (sensor steps and contralateral steps); red squares indicate HS events, whereas red circles indicate TO events—both events are detected only for the leg which is carrying the sensor. In (b) the estimated parameters are shown: shaded areas highlight stance periods.

Figure 8

Bland–Altman plots—Trunk vs. Shoe measurements.

Figure 9

Bland–Altman plots—Pocket vs. Shoe measurements.