Estimation of horizontal running power using foot-worn inertial measurement units
Feedback of power during running is a promising tool for training and determining pacing strategies. However, current power estimation methods show low validity and are not customized for running on different slopes. To address this issue, we developed three machine-learning models to estimate peak...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2023-06-01
|
| Series: | Frontiers in Bioengineering and Biotechnology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fbioe.2023.1167816/full |
| _version_ | 1797797765332336640 |
|---|---|
| author | Salil Apte Mathieu Falbriard Frédéric Meyer Frédéric Meyer Grégoire P. Millet Vincent Gremeaux Vincent Gremeaux Kamiar Aminian |
| author_facet | Salil Apte Mathieu Falbriard Frédéric Meyer Frédéric Meyer Grégoire P. Millet Vincent Gremeaux Vincent Gremeaux Kamiar Aminian |
| author_sort | Salil Apte |
| collection | DOAJ |
| description | Feedback of power during running is a promising tool for training and determining pacing strategies. However, current power estimation methods show low validity and are not customized for running on different slopes. To address this issue, we developed three machine-learning models to estimate peak horizontal power for level, uphill, and downhill running using gait spatiotemporal parameters, accelerometer, and gyroscope signals extracted from foot-worn IMUs. The prediction was compared to reference horizontal power obtained during running on a treadmill with an embedded force plate. For each model, we trained an elastic net and a neural network and validated it with a dataset of 34 active adults across a range of speeds and slopes. For the uphill and level running, the concentric phase of the gait cycle was considered, and the neural network model led to the lowest error (median ± interquartile range) of 1.7% ± 12.5% and 3.2% ± 13.4%, respectively. The eccentric phase was considered relevant for downhill running, wherein the elastic net model provided the lowest error of 1.8% ± 14.1%. Results showed a similar performance across a range of different speed/slope running conditions. The findings highlighted the potential of using interpretable biomechanical features in machine learning models for the estimating horizontal power. The simplicity of the models makes them suitable for implementation on embedded systems with limited processing and energy storage capacity. The proposed method meets the requirements for applications needing accurate near real-time feedback and complements existing gait analysis algorithms based on foot-worn IMUs. |
| first_indexed | 2024-03-13T03:53:18Z |
| format | Article |
| id | doaj.art-dde8071f5b2747ab9d32b75d966660be |
| institution | Directory Open Access Journal |
| issn | 2296-4185 |
| language | English |
| last_indexed | 2024-03-13T03:53:18Z |
| publishDate | 2023-06-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Bioengineering and Biotechnology |
| spelling | doaj.art-dde8071f5b2747ab9d32b75d966660be2023-06-22T09:28:50ZengFrontiers Media S.A.Frontiers in Bioengineering and Biotechnology2296-41852023-06-011110.3389/fbioe.2023.11678161167816Estimation of horizontal running power using foot-worn inertial measurement unitsSalil Apte0Mathieu Falbriard1Frédéric Meyer2Frédéric Meyer3Grégoire P. Millet4Vincent Gremeaux5Vincent Gremeaux6Kamiar Aminian7Laboratory of Movement Analysis and Measurement, École Polytechnique Fédérale de Lausanne, Lausanne, SwitzerlandLaboratory of Movement Analysis and Measurement, École Polytechnique Fédérale de Lausanne, Lausanne, SwitzerlandDigital Signal Processing Group, Department of Informatics, University of Oslo, Oslo, NorwayInstitute of Sport Sciences, University of Lausanne, Lausanne, SwitzerlandInstitute of Sport Sciences, University of Lausanne, Lausanne, SwitzerlandInstitute of Sport Sciences, University of Lausanne, Lausanne, SwitzerlandSport Medicine Unit, Division of Physical Medicine and Rehabilitation, Swiss Olympic Medical Center, Lausanne University Hospital, Lausanne, SwitzerlandLaboratory of Movement Analysis and Measurement, École Polytechnique Fédérale de Lausanne, Lausanne, SwitzerlandFeedback of power during running is a promising tool for training and determining pacing strategies. However, current power estimation methods show low validity and are not customized for running on different slopes. To address this issue, we developed three machine-learning models to estimate peak horizontal power for level, uphill, and downhill running using gait spatiotemporal parameters, accelerometer, and gyroscope signals extracted from foot-worn IMUs. The prediction was compared to reference horizontal power obtained during running on a treadmill with an embedded force plate. For each model, we trained an elastic net and a neural network and validated it with a dataset of 34 active adults across a range of speeds and slopes. For the uphill and level running, the concentric phase of the gait cycle was considered, and the neural network model led to the lowest error (median ± interquartile range) of 1.7% ± 12.5% and 3.2% ± 13.4%, respectively. The eccentric phase was considered relevant for downhill running, wherein the elastic net model provided the lowest error of 1.8% ± 14.1%. Results showed a similar performance across a range of different speed/slope running conditions. The findings highlighted the potential of using interpretable biomechanical features in machine learning models for the estimating horizontal power. The simplicity of the models makes them suitable for implementation on embedded systems with limited processing and energy storage capacity. The proposed method meets the requirements for applications needing accurate near real-time feedback and complements existing gait analysis algorithms based on foot-worn IMUs.https://www.frontiersin.org/articles/10.3389/fbioe.2023.1167816/fullbiomechanicsmachine learningwearable sensorsmovement analysissignal processingquantitative feedback |
| spellingShingle | Salil Apte Mathieu Falbriard Frédéric Meyer Frédéric Meyer Grégoire P. Millet Vincent Gremeaux Vincent Gremeaux Kamiar Aminian Estimation of horizontal running power using foot-worn inertial measurement units Frontiers in Bioengineering and Biotechnology biomechanics machine learning wearable sensors movement analysis signal processing quantitative feedback |
| title | Estimation of horizontal running power using foot-worn inertial measurement units |
| title_full | Estimation of horizontal running power using foot-worn inertial measurement units |
| title_fullStr | Estimation of horizontal running power using foot-worn inertial measurement units |
| title_full_unstemmed | Estimation of horizontal running power using foot-worn inertial measurement units |
| title_short | Estimation of horizontal running power using foot-worn inertial measurement units |
| title_sort | estimation of horizontal running power using foot worn inertial measurement units |
| topic | biomechanics machine learning wearable sensors movement analysis signal processing quantitative feedback |
| url | https://www.frontiersin.org/articles/10.3389/fbioe.2023.1167816/full |
| work_keys_str_mv | AT salilapte estimationofhorizontalrunningpowerusingfootworninertialmeasurementunits AT mathieufalbriard estimationofhorizontalrunningpowerusingfootworninertialmeasurementunits AT fredericmeyer estimationofhorizontalrunningpowerusingfootworninertialmeasurementunits AT fredericmeyer estimationofhorizontalrunningpowerusingfootworninertialmeasurementunits AT gregoirepmillet estimationofhorizontalrunningpowerusingfootworninertialmeasurementunits AT vincentgremeaux estimationofhorizontalrunningpowerusingfootworninertialmeasurementunits AT vincentgremeaux estimationofhorizontalrunningpowerusingfootworninertialmeasurementunits AT kamiaraminian estimationofhorizontalrunningpowerusingfootworninertialmeasurementunits |