Indoor running temporal variability for different running speeds, treadmill inclinations, and three different estimation strategies

PLOS ONE RESEARCH ARTICLE Indoor running temporal variability for different running speeds, treadmill inclinations, and three different estimation strategies Andrea Zignoli ID1*, Antoine Godin2, Laurent Mourot ID2 1 Department of Industrial Engineering, University of Trento, Trento, Italy, 2 Prognostic Factors and Regulatory Factors of Cardiac and Vascular Pathologies (EA3920), Exercise Performance Health Innovation (EPHI) platform, University of Franche-Comté, Besançon, France a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Zignoli A, Godin A, Mourot L (2023) Indoor running temporal variability for different running speeds, treadmill inclinations, and three different estimation strategies. PLoS ONE 18(7): e0287978. https://doi.org/10.1371/journal. pone.0287978 Editor: Alessandro Mengarelli, Università Politecnica delle Marche Facoltà di Ingegneria: Universita Politecnica delle Marche Facolta di Ingegneria, ITALY Received: March 13, 2023 Accepted: June 19, 2023 Published: July 20, 2023 Peer Review History: PLOS recognizes the benefits of transparency in the peer review process; therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal.pone.0287978 * Abstract Inertial measurement units (IMU) constitute a light and cost-effective alternative to goldstandard measurement systems in the assessment of running temporal variables. IMU data collected on 20 runners running at different speeds (80, 90, 100, 110 and 120% of preferred running speed) and treadmill inclination (±2, ±5, and ±8%) were used here to predict the following temporal variables: stride frequency, duty factor, and two indices of running variability such as the detrended fluctuation analysis alpha (DFA-α) and the Higuchi’s D (HG-D). Three different estimation methodologies were compared: 1) a gold-standard optoelectronic device (which provided the reference values), 2) IMU placed on the runner’s feet, 3) a single IMU on the runner’s thorax used in conjunction with a machine learning algorithm with a short 2-second or a long 120-second window as input. A two-way ANOVA was used to test the presence of significant (p<0.05) differences due to the running condition or to the estimation methodology. The findings of this study suggest that using both IMU configurations for estimating stride frequency can be effective and comparable to the gold-standard. Additionally, the results indicate that the use of a single IMU on the thorax with a machine learning algorithm can lead to more accurate estimates of duty factor than the strategy of the IMU on the feet. However, caution should be exercised when using these techniques to measure running variability indices. Estimating DFA-α from a short 2-second time window was possible only in level running but not in downhill running and it could not accurately estimate HGD across all running conditions. By taking a long 120-second window a machine learning algorithm could improve the accuracy in the estimation of DFA-α in all running conditions. By taking these factors into account, researchers and practitioners can make informed decisions about the use of IMU technology in measuring running biomechanics. Copyright: © 2023 Zignoli et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. PLOS ONE | https://doi.org/10.1371/journal.pone.0287978 July 20, 2023 1 / 21 PLOS ONE Data Availability Statement: Data samples will be uploaded at https://www.kaggle.com/datasets/ andreazignoli/prissiv Funding: This project was partially funded by ANTA with the ANTA Sports Award. The award has been presented at the 25th Annual Congress of the European College of Sport Science, in July 2020 in Seville, Spain. The prize has been awarded on the scientific merit of the proposal and to "stimulate research on next-generation sports science technology: uncovering insights in sports performance and injury prevention." The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. Indoor running temporal variability Introduction Running temporal variables such as contact time, flight time, and duty cycle, are widely adopted in the study of the running biomechanics [1], and they are used to assess and characterize different running conditions [2]. Running structure variability [3], or simply ’variability’ in the context of this manuscript, refers to a self-similarity assessment -computed on a strideby-stride basis- of the running temporal variables. Two running variability indices of interest are the detrended fluctuation analysis long-range α (DFA-α) and the Higuchi’s D (HG-D) [4]. Particularly, DFA-α is a measure of the correlation properties of a signal, and it is related to the size of fluctuations in signal changes with the length of time over which the fluctuations are measured. HG-D is a measure of the fractal dimension of a signal, which is a way of characterizing its overall complexity. Variability is an intrinsic property of any temporal signal. However, a common practice in biomechanics is to assign to a locomotion pattern the characteristics of a signal with equal variability, e.g.: 1) DFA-α values closer to 1 (HG-D!1.8) are characteristics of the pink-noise, and hence of a flexible-and-adaptable locomotor behavior; 2) DFA-α values closer to 0.5 (HG-D!2) are characteristics of the white-noise, and hence of an impaired locomotion pattern which is not able to adapt to the environment; 3) DFA-α values closer to 0.75 (HG-D!1.9) indicate a locomotion pattern with both characteristics of structure and functional variance. These analogies have been widely adopted irrespectively from the locomotion strategy, e.g.: both in walking [5–8] and running [9–12]. Quite interestingly, Jordan et al. [9, 10] found a U-shaped relationship between running speed and DFA-α values computed with stride interval time. The minimum value of this relationship was found at the preferred running speed with a correspondent DFA-α value very close to 0.8. Decreasing variability has been observed after a heavy period of training [12] and in runners previously affected by a running-related injury [13]. The idea of monitoring variability to flag impaired running patterns potentially prone to injuries is therefore very appealing, but counts only few actual attempts [14–16]. To correctly assess running variability, accurate running temporal variables must be accurately estimated. Running variability is indeed best assessed in laboratory conditions, where optoelectronic devices, force platforms, motio (...truncated)