Automated event detection algorithms in pathological gait

Bruening, Dustin A.; Ridge, Sarah T.

doi:10.1016/j.gaitpost.2013.08.023

Cited by 64 publications

(91 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Bruning and Ridge recently showed that algorithms that similarly depend on horizontal foot positions are more accurate in CPP patients with horizontal foot patterns, found in equinus or sliding movements, and are inaccurate in more vertical foot patterns as found in steppage gait [10]. Since the CPP cohort trials were not subdivided into different gait patterns, it may be possible that the PHMD algorithm had varied accuracy, depending on the patients' step patterns, similar to the HPA algorithm.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Automatic initial contact detection during overground walking for clinical use

et al. 2014

View full text Add to dashboard Cite

Section: Discussionmentioning

confidence: 99%

“…After filtering, upper and lower envelopes will be calculated as cubic Bezier spline approximation from all extremes using the same techniques as Hilbert-Huang transformation [9] and empirical mode decomposition [10]. They are used to select the areas where an IC event occurs.…”

Section: Step 3-calculation Of Dynamic Limitsmentioning

confidence: 99%

Automatic initial contact detection during overground walking for clinical use

et al. 2014

View full text Add to dashboard Cite

“…Thus, identifying these events is the key to many gait analysis applications [2]- [9] that would benefit from long-term, continuous monitoring in humans' natural environment, enabling gait assessment and interventions that have not previously been possible [10]. The present state of practice is to perform clinical gait analysis in controlled gait labs equipped with stationary sensor systems such as motion capture systems and force plates [11]. Although these systems provide rich and accurate information, they are inadequate for use in daily life as they…”

Section: Introductionmentioning

confidence: 99%

Gait Event Detection in Real-World Environment for Long-Term Applications: Incorporating Domain Knowledge Into Time-Frequency Analysis

Khandelwal

Wickström

2016

IEEE Trans. Neural Syst. Rehabil. Eng.

View full text Add to dashboard Cite

PostprintThis is the accepted version of a paper published in IEEE transactions on neural systems and rehabilitation engineering. This paper has been peer-reviewed but does not include the final publisher proof-corrections or journal pagination. This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Abstract-Detecting gait events is the key to many gait analysis applications that would benefit from continuous monitoring or long-term analysis. Most gait event detection algorithms using wearable sensors that offer a potential for use in daily living have been developed from data collected in controlled indoor experiments. However, for real-word applications, it is essential that the analysis is carried out in humans' natural environment; that involves different gait speeds, changing walking terrains, varying surface inclinations and regular turns among other factors. Existing domain knowledge in the form of principles or underlying fundamental gait relationships can be utilized to drive and support the data analysis in order to develop robust algorithms that can tackle real-world challenges in gait analysis. This paper presents a novel approach that exhibits how domain knowledge about human gait can be incorporated into time-frequency analysis to detect gait events from longterm accelerometer signals. The accuracy and robustness of the proposed algorithm are validated by experiments done in indoor and outdoor environments with approximately 93,600 gait events in total. The proposed algorithm exhibits consistently high performance scores across all datasets in both, indoor and outdoor environments.

show abstract

“…Heel-strike and toe-off events were determined using four kinematic-based algorithms: (1) horizontal heel and toe position relative to the pelvis (Zeni et al, 2008); (2) absolute horizontal heel and toe velocity (Zeni et al, 2008); (3) vertical velocity of the foot center-of-mass (O'Connor et al, 2007); and (4) sagittal resultant velocity of the heel and toe markers (Ghoussayni et al, 2004) using automated detection thresholds adjusted for walking speed (Bruening and Ridge, 2014). These algorithms were chosen based on existing evidence of good reliability for treadmill walking, and to an extent, their robustness in potential applicability to pathological gait (Bruening and Ridge, 2014).…”

Section: Analysesmentioning

confidence: 99%

“…These algorithms were chosen based on existing evidence of good reliability for treadmill walking, and to an extent, their robustness in potential applicability to pathological gait (Bruening and Ridge, 2014). For purposes of comparing the aforementioned kinematic-based methods to a "gold standard", GRF-based events were also determined using a 20 N ascending and descending threshold in vertical force component, adjusted upward slightly from previous thresholds (10 N) to account for additional noise from the treadmill belts/platform (Tirosh and Sparrow, 2003;Zeni et al, 2008).…”

Section: Analysesmentioning

confidence: 99%

A comparison of kinematic-based gait event detection methods in a self-paced treadmill application

Hendershot

Mahon

Pruziner

2016

Journal of Biomechanics

View full text Add to dashboard Cite

Kinematic-based algorithms for detecting gait events are efficient and useful in the absence of (reliable) kinetic data. However, the validity of these kinematic-based algorithms for self-paced treadmill walking is unknown, particularly given the influence of walking speed on such data. We quantified offsets in event detection of four foot kinematics-based algorithms (horizontal position, horizontal velocity, vertical velocity, and sagittal resultant velocity) relative to events determined by a threshold in vertical ground reaction force among seven uninjured individuals - and nine with unilateral transtibial amputation - walking on a self-paced treadmill. Across walking speeds from 0.48-1.64m/s (0.5-31.7% CV), offsets ranged from -7 to +3 frames (≈83.3ms) in heel strike, and -3 to +5 frames (≈66.6 ms) in toe off. Regardless of method, offsets in heel strike were not influenced (-0.010.61) by variability in walking speed. However, offsets in toe-off were positively correlated with variability in walking speed for the horizontal position (r=0.539; P<0.001) and velocity (r=0.463; P<0.001) algorithms, and negatively correlated (r=-0.317; P<0.001) for the vertical velocity algorithm; offsets from the sagittal resultant velocity algorithm, with thresholds adjusted for walking speed, were not strongly associated (r=0.126; P=0.27). Although relatively minimal offsets support the applicability of these algorithms to self-paced walking, for individuals with asymptomatic and pathological gait patterns, sagittal resultant velocity of the foot produces the most consistent event detection over the widest range of (and variability in) walking speeds.

show abstract

Automated event detection algorithms in pathological gait

Cited by 64 publications

References 16 publications

Automatic initial contact detection during overground walking for clinical use

Automatic initial contact detection during overground walking for clinical use

Gait Event Detection in Real-World Environment for Long-Term Applications: Incorporating Domain Knowledge Into Time-Frequency Analysis

A comparison of kinematic-based gait event detection methods in a self-paced treadmill application

Contact Info

Product

Resources

About