Non steady-state prediction of the instantaneous metabolic energy expenditure in pathological gait: an exploratory study

Non steady-state prediction of the instantaneous metabolic energy expenditure in pathological gait: an exploratory study

Filius, Suzanne (TU Delft Mechanical, Maritime and Materials Engineering)

Harlaar, J. (mentor)
Dodou, D. (graduation committee)
Vallery, H. (graduation committee)
de Winter, J.C.F. (graduation committee)

Delft University of Technology

Biomedical Engineering

2019-09-05

BACKGROUND - In a variety of patients with locomotive disorders gait efficiency is often assessed using mobile metabolic gas analysis during rehabilitative interventions. Conventionally, these measurements take 6 minutes per evaluated condition (e.g. with/without ankle-foot-orthosis; AFO). This duration is required since there is a dynamical delay between the instantaneous metabolic energy expenditure (ImEE), and the respiratory response measured at the mouth. Moreover, the breath-by-breath data is sparely sampled and noisy. Gait efficiency is therefore computed from an averaged (i.e. of 1-3 minutes) respiratory response during a steady-state metabolism (i.e. reached after 3-4 minutes). This is time consuming and can be exhausting for patients with severe gait impairments. In up-coming human-in-the-loop techniques, fast (2-3 minutes) predictions of the ImEE are made using an Instantaneous Cost Mapping (ICM) model. The ICM is based on a first-order response with a known time constant (τ) to a change in an external load (e.g. another walking speed). The τ is either identified on a subject-specific basis or set to a fixed average time constant already identified for healthy adults (42s; Selinger & Donelan, 2014). It is unknown whether this technique could be applied in patient populations with additional gait impairments. When applicable, the ICM model could speed-up the assessment of the gait efficiency and open ways to make human-in-the-loop protocols feasible device optimization (e.g. AFO tuning) in rehabilitation. AIM - This study investigates whether the subject-specific and/or general ICM model could reduce the required measurement duration in a variety of patients that cope with gait impairments as consequence of a neurology or neuromuscular disease. Secondarily, it explores whether there are differences in the identified subject-specific τ among the pathologies.METHODS - Post hoc metabolic data, recorded in the period of 2006 to 2019 within the Amsterdam University Medical Centers (UMC) was collected, containing walking trials of patients with Multiple Sclerosis, Cerebral Palsy, several neuromuscular disorders, healthy adults and typically developing children. The ICM model was tested for the subject-specific and general τ to estimate the ImEE. First, using full measurement durations and later the identified reduced measurement durations. The model outcome was assessed for individually correctness. RESULTS - Results show (n = 28) that the ICM model correctly estimates the ImEE using the total conventional measurement duration (subject-specific: 96%; general: 89% individually correctness). For the identified reduced measure duration (subject-specific: 2m33s; general: 3m:04s) reduced the individually correctness to 57%. No differences among the groups were found for the subject-specific τ and the average (τ = 41.7s ± 13.5s) was similar to the reported τ for healthy adults. CONCLUSIONS - Based on the results, it can be concluded that the ICM model show similar results to healthy adults from literature. This offers prospect for the clinical application of the model in rehabilitation depending on the purposes of use. Clinicians should consider the balance between the individually correctness and measurement duration of the predictions. Results should be treated with caution due to the small and heterogeneous sample size. Future research is needed to identify differences in subject-specific τ among pathologies.

human-in-the-loop
instantaneous-cost-mapping
wearable device optimization
metabolic gas analysis
non-steady-state prediction
gait impairments

http://resolver.tudelft.nl/uuid:da503201-95b8-41fe-99b9-58d227bd73be

Student theses

master thesis

© 2019 Suzanne Filius