Deep Learning Unravels Differences Between Kinematic and Kinetic Gait Cycle Time Series from Two Control Samples of Healthy Children Assessed in Two Different Gait Laboratories
We investigate the application of deep learning in comparing gait cycle time series from two groups of healthy children, each assessed in different gait laboratories. Both laboratories used similar gait analysis protocols with minimal differences in data collection. Utilizing a ResNet-based deep lea...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/25/1/110 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841548918725279744 |
|---|---|
| author | Alfonso de Gorostegui Damien Kiernan Juan-Andrés Martín-Gonzalo Javier López-López Irene Pulido-Valdeolivas Estrella Rausell Massimiliano Zanin David Gómez-Andrés |
| author_facet | Alfonso de Gorostegui Damien Kiernan Juan-Andrés Martín-Gonzalo Javier López-López Irene Pulido-Valdeolivas Estrella Rausell Massimiliano Zanin David Gómez-Andrés |
| author_sort | Alfonso de Gorostegui |
| collection | DOAJ |
| description | We investigate the application of deep learning in comparing gait cycle time series from two groups of healthy children, each assessed in different gait laboratories. Both laboratories used similar gait analysis protocols with minimal differences in data collection. Utilizing a ResNet-based deep learning model, we successfully identified the source laboratory of each dataset, achieving a high classification accuracy across multiple gait parameters. To address the inter-laboratory differences, we explored various pre-processing methods and time series properties that may have been detected by the algorithm. We found that the standardization of the time series values was a successful approach to decrease the ability of the model to distinguish between the two centers. Our findings also reveal that differences in the power spectra and autocorrelation structures of the datasets play a significant role in the model performance. Our study emphasizes the importance of standardized protocols and robust data pre-processing to enhance the transferability of machine learning models across clinical settings, particularly for deep learning approaches. |
| format | Article |
| id | doaj-art-551fe2229bce4c4a88074f323861423c |
| institution | Kabale University |
| issn | 1424-8220 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj-art-551fe2229bce4c4a88074f323861423c2025-01-10T13:20:54ZengMDPI AGSensors1424-82202024-12-0125111010.3390/s25010110Deep Learning Unravels Differences Between Kinematic and Kinetic Gait Cycle Time Series from Two Control Samples of Healthy Children Assessed in Two Different Gait LaboratoriesAlfonso de Gorostegui0Damien Kiernan1Juan-Andrés Martín-Gonzalo2Javier López-López3Irene Pulido-Valdeolivas4Estrella Rausell5Massimiliano Zanin6David Gómez-Andrés7PhD Program in Neuroscience, Universidad Autonoma de Madrid-Cajal Institute, 28029 Madrid, SpainMovement Analysis Laboratory, Central Remedial Clinic, Clontarf, D03 R973 Dublin, IrelandEscuela Universitaria de Fisioterapia de la ONCE, Universidad Autónoma de Madrid, 28034 Madrid, SpainDepartment of Rehabilitation, Hospital Universitario Infanta Sofía, Fundación para la Investigación e Innovación Biomédica del Hospital Universitario Infanta Sofía y Hospital del Henares, San Sebastián de los Reyes, 28702 Madrid, SpainDepartment of Anatomy, Histology & Neuroscience, School of Medicine, Universidad Autónoma de Madrid (UAM), 28029 Madrid, SpainDepartment of Anatomy, Histology & Neuroscience, School of Medicine, Universidad Autónoma de Madrid (UAM), 28029 Madrid, SpainInstituto de Física Interdisciplinar y Sistemas Complejos IFISC (CSIC-UIB), Campus UIB, 07122 Palma de Mallorca, SpainPediatric Neurology, ERN-RND, Euro-NMD, Vall d’Hebron Institut de Recerca (VHIR), Hospital Universitari Vall d’Hebron, Vall d’Hebron Barcelona Hospital Campus, Passeig Vall d’Hebron 119-129, 08035 Barcelona, SpainWe investigate the application of deep learning in comparing gait cycle time series from two groups of healthy children, each assessed in different gait laboratories. Both laboratories used similar gait analysis protocols with minimal differences in data collection. Utilizing a ResNet-based deep learning model, we successfully identified the source laboratory of each dataset, achieving a high classification accuracy across multiple gait parameters. To address the inter-laboratory differences, we explored various pre-processing methods and time series properties that may have been detected by the algorithm. We found that the standardization of the time series values was a successful approach to decrease the ability of the model to distinguish between the two centers. Our findings also reveal that differences in the power spectra and autocorrelation structures of the datasets play a significant role in the model performance. Our study emphasizes the importance of standardized protocols and robust data pre-processing to enhance the transferability of machine learning models across clinical settings, particularly for deep learning approaches.https://www.mdpi.com/1424-8220/25/1/110gaitdeep learningexternal validitychildren |
| spellingShingle | Alfonso de Gorostegui Damien Kiernan Juan-Andrés Martín-Gonzalo Javier López-López Irene Pulido-Valdeolivas Estrella Rausell Massimiliano Zanin David Gómez-Andrés Deep Learning Unravels Differences Between Kinematic and Kinetic Gait Cycle Time Series from Two Control Samples of Healthy Children Assessed in Two Different Gait Laboratories Sensors gait deep learning external validity children |
| title | Deep Learning Unravels Differences Between Kinematic and Kinetic Gait Cycle Time Series from Two Control Samples of Healthy Children Assessed in Two Different Gait Laboratories |
| title_full | Deep Learning Unravels Differences Between Kinematic and Kinetic Gait Cycle Time Series from Two Control Samples of Healthy Children Assessed in Two Different Gait Laboratories |
| title_fullStr | Deep Learning Unravels Differences Between Kinematic and Kinetic Gait Cycle Time Series from Two Control Samples of Healthy Children Assessed in Two Different Gait Laboratories |
| title_full_unstemmed | Deep Learning Unravels Differences Between Kinematic and Kinetic Gait Cycle Time Series from Two Control Samples of Healthy Children Assessed in Two Different Gait Laboratories |
| title_short | Deep Learning Unravels Differences Between Kinematic and Kinetic Gait Cycle Time Series from Two Control Samples of Healthy Children Assessed in Two Different Gait Laboratories |
| title_sort | deep learning unravels differences between kinematic and kinetic gait cycle time series from two control samples of healthy children assessed in two different gait laboratories |
| topic | gait deep learning external validity children |
| url | https://www.mdpi.com/1424-8220/25/1/110 |
| work_keys_str_mv | AT alfonsodegorostegui deeplearningunravelsdifferencesbetweenkinematicandkineticgaitcycletimeseriesfromtwocontrolsamplesofhealthychildrenassessedintwodifferentgaitlaboratories AT damienkiernan deeplearningunravelsdifferencesbetweenkinematicandkineticgaitcycletimeseriesfromtwocontrolsamplesofhealthychildrenassessedintwodifferentgaitlaboratories AT juanandresmartingonzalo deeplearningunravelsdifferencesbetweenkinematicandkineticgaitcycletimeseriesfromtwocontrolsamplesofhealthychildrenassessedintwodifferentgaitlaboratories AT javierlopezlopez deeplearningunravelsdifferencesbetweenkinematicandkineticgaitcycletimeseriesfromtwocontrolsamplesofhealthychildrenassessedintwodifferentgaitlaboratories AT irenepulidovaldeolivas deeplearningunravelsdifferencesbetweenkinematicandkineticgaitcycletimeseriesfromtwocontrolsamplesofhealthychildrenassessedintwodifferentgaitlaboratories AT estrellarausell deeplearningunravelsdifferencesbetweenkinematicandkineticgaitcycletimeseriesfromtwocontrolsamplesofhealthychildrenassessedintwodifferentgaitlaboratories AT massimilianozanin deeplearningunravelsdifferencesbetweenkinematicandkineticgaitcycletimeseriesfromtwocontrolsamplesofhealthychildrenassessedintwodifferentgaitlaboratories AT davidgomezandres deeplearningunravelsdifferencesbetweenkinematicandkineticgaitcycletimeseriesfromtwocontrolsamplesofhealthychildrenassessedintwodifferentgaitlaboratories |