Standardizing approaches on automatic gait event detection for use in pediatric pathological populations - A multicenter evaluation

Visscher, Rosa M. S.; Armand, Stéphane; Heitzmann, Daniel; Wolf, Sebastian I.; Lempereur, Mathieu; Leboeuf, Fabien; Grant-Beuttler, Marybeth; Soangra, Rahul; Schwartz, Michael; Claes, Claudia; Galli, Manuela; Camuncoli, Federica; Gwerder, Michelle; Taylor, William R.; Brunner, Reinald; Singh, Navrag B.; Viehweger, Elke

doi:10.1016/j.gaitpost.2021.09.155

Cited by 1 publication

(6 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Minor differences in the general setup or marker placement may introduce a bias learned by the event detection algorithm, affecting the accuracy of the algorithm when applied to data from different laboratories. Visscher et al [17] showed this for heuristic approaches, where a difference of up to 15 ms occurred between laboratories. Smaller differences in results were shown in laboratories with a more comparable setup [17].…”

Section: Introductionmentioning

confidence: 89%

“…Visscher et al [17] showed this for heuristic approaches, where a difference of up to 15 ms occurred between laboratories. Smaller differences in results were shown in laboratories with a more comparable setup [17]. Therefore, it is important to understand how a model performs between different gait laboratories.…”

Section: Introductionmentioning

confidence: 89%

“…With an absolute error of 10 ms for IC and 17.5 ms for FO estimation between different gait laboratories this bias could also translate to deep learning-based approaches as it is comparable to the difference of 10.2 ms for the IC between the DeepEvent-original and DeepEvent-retrained models in our results. Laboratories with similar setups and marker placement protocols show higher agreement in the results [17]. Therefore, Visscher et al [17] suggested the need for a multi-center approach to improve reproducibility and validity.…”

Section: Generalizability Across Laboratoriesmentioning

confidence: 99%

Section: Generalizability Across Laboratoriesmentioning

confidence: 99%

“…Furthermore, the marker placement and setup differences between laboratories as well as the manually set gait cycle events by professional clinicians could also have added a certain bias as reported in [17]. To date, only the heuristic approach of Ghoussayni et al [5] was compared between nine different gait laboratories [17]. With an absolute error of 10 ms for IC and 17.5 ms for FO estimation between different gait laboratories this bias could also translate to deep learning-based approaches as it is comparable to the difference of 10.2 ms for the IC between the DeepEvent-original and DeepEvent-retrained models in our results.…”

Section: Generalizability Across Laboratoriesmentioning

confidence: 99%

See 4 more Smart Citations

Robust deep learning-based gait event detection across various pathologies

et al. 2023

View full text Add to dashboard Cite

The correct estimation of gait events is essential for the interpretation and calculation of 3D gait analysis (3DGA) data. Depending on the severity of the underlying pathology and the availability of force plates, gait events can be set either manually by trained clinicians or detected by automated event detection algorithms. The downside of manually estimated events is the tedious and time-intensive work which leads to subjective assessments. For automated event detection algorithms, the drawback is, that there is no standardized method available. Algorithms show varying robustness and accuracy on different pathologies and are often dependent on setup or pathology-specific thresholds. In this paper, we aim at closing this gap by introducing a novel deep learning-based gait event detection algorithm called IntellEvent, which shows to be accurate and robust across multiple pathologies. For this study, we utilized a retrospective clinical 3DGA dataset of 1211 patients with four different pathologies (malrotation deformities of the lower limbs, club foot, infantile cerebral palsy (ICP), and ICP with only drop foot characteristics) and 61 healthy controls. We propose a recurrent neural network architecture based on long-short term memory (LSTM) and trained it with 3D position and velocity information to predict initial contact (IC) and foot off (FO) events. We compared IntellEvent to a state-of-the-art heuristic approach and a machine learning method called DeepEvent. IntellEvent outperforms both methods and detects IC events on average within 5.4 ms and FO events within 11.3 ms with a detection rate of ≥ 99% and ≥ 95%, respectively. Our investigation on generalizability across laboratories suggests that models trained on data from a different laboratory need to be applied with care due to setup variations or differences in capturing frequencies.

show abstract

Section: Introductionmentioning

confidence: 89%

Section: Introductionmentioning

confidence: 89%