An Automatic Gait Analysis Pipeline for Wearable Sensors: A Pilot Study in Parkinson’s Disease

Abstract

The use of wearable sensors allows continuous recordings of physical activity from participants in free-living or at-home clinical studies. The large amount of data collected demands automatic analysis pipelines to extract gait parameters that can be used as clinical endpoints. We introduce a deep learning-based automatic pipeline for wearables that processes tri-axial accelerometry data and extracts gait events—bout segmentation, initial contact (IC), and final contact (FC)—from a single sensor located at either the lower back (near L5), shin or wrist. The gait events detected are posteriorly used for gait parameter estimation, such as step time, length, and symmetry. We report results from a leave-one-subject-out (LOSO) validation on a pilot study dataset of five participants clinically diagnosed with Parkinson’s disease (PD) and six healthy controls (HC). Participants wore sensors at three body locations and walked on a pressure-sensing walkway to obtain reference gait data. Mean absolute errors (MAE) for the IC events ranged from 22.82 to 33.09 milliseconds (msecs) for the lower back sensor while for the shin and wrist sensors, MAE ranges were 28.56–64.66 and 40.19–72.50 msecs, respectively. For the FC-event detection, MAE ranges were 29.06–48.42, 40.19–72.70 and 36.06–60.18 msecs for the lumbar, wrist and shin sensors, respectively. Intraclass correlation coefficients, ICC(2,k), between the estimated parameters and the reference data resulted in good-to-excellent agreement (ICC ≥ 0.84) for the lumbar and shin sensors, excluding the double support time (ICC = 0.37 lumbar and 0.38 shin) and swing time (ICC = 0.55 lumbar and 0.59 shin). The wrist sensor also showed good agreements, but the ICCs were lower overall than for the other two sensors. Our proposed analysis pipeline has the potential to extract up to 100 gait-related parameters, and we expect our contribution will further support developments in the fields of wearable sensors, digital health, and remote monitoring in clinical trials.

Validity and Reliability of a Smartphone App for Gait and Balance Assessment

“…A possible explanation for these results is the signal noise caused by the integration procedure applied as part of the inverted pendulum gait model used for obtaining these parameters from acceleration time series [ 43 ]. Analysis of gait data with neural networks may lead to more promising results [ 66 ].…”

Validity and reliability of the Apple Health app on iPhone for measuring gait parameters in children, adults, and seniors

“…Agreement of the OneStep app with different reference standards (APDM Mobility Lab, Zeno walkway) was higher for gait speed (ICC = 0.94, Pearson correlation coefficient [ r ] = 0.89–0.91) and step length (ICC = 0.80, r = 0.65–0.84) than for double limb support (ICC = 0.52, r = 0.61– 0 .62). The lower validity for measuring double support time has also been reported for stand-alone IMUs in normal 68 and pathological gait 69 , 70 . As previously discussed, the difficulty of accurately detecting both heel strike and toe-off for estimating this gait parameter has often been mentioned in this context as a potential reason for greater measurement errors with this gait parameter 65 – 67 , which may also be the case for the Health app.…”

Wearable sensor devices can automatically identify the ON-OFF status of patients with Parkinson's disease through an interpretable machine learning model

“…Some studies also suggested that the waist and upper limbs were the optimal locations for collecting kinematic parameters in patients with PD ( 41 , 52 – 54 ). Peraza et al ( 55 ) proposed an automatic gait analysis process based on DL algorithms, with data sourced from triaxial accelerometers placed on the lower limbs, trunk (waist), and upper limbs. The results showed that data from single triaxial accelerometers on the lower limbs and trunk (waist) performed better than those from the upper limbs in assessing gait in patients with PD and healthy elderly people ( 55 ).…”