Explaining the unique nature of individual gait patterns with deep learning
- Publikationstyp:
- Zeitschriftenaufsatz
- Metadaten:
-
- Autoren
- Fabian Horst
- Sebastian Lapuschkin
- Wojciech Samek
- Klaus-Robert Mueller
- Wolfgang I Schoellhorn
- Autoren-URL
- https://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=fis-test-1&SrcAuth=WosAPI&KeyUT=WOS:000459094800072&DestLinkType=FullRecord&DestApp=WOS_CPL
- DOI
- 10.1038/s41598-019-38748-8
- Externe Identifier
- Clarivate Analytics Document Solution ID: HL9WW
- PubMed Identifier: 30787319
- ISSN
- 2045-2322
- Zeitschrift
- SCIENTIFIC REPORTS
- Artikelnummer
- ARTN 2391
- Datum der Veröffentlichung
- 2019
- Status
- Published
- Titel
- Explaining the unique nature of individual gait patterns with deep learning
- Sub types
- Article
- Ausgabe der Zeitschrift
- 9
Datenquelle: Web of Science (Lite)
- Andere Metadatenquellen:
-
- Abstract
- <jats:title>Abstract</jats:title><jats:p>Machine learning (ML) techniques such as (deep) artificial neural networks (DNN) are solving very successfully a plethora of tasks and provide new predictive models for complex physical, chemical, biological and social systems. However, in most cases this comes with the disadvantage of acting as a black box, rarely providing information about what made them arrive at a particular prediction. This black box aspect of ML techniques can be problematic especially in medical diagnoses, so far hampering a clinical acceptance. The present paper studies the uniqueness of individual gait patterns in clinical biomechanics using DNNs. By attributing portions of the model predictions back to the input variables (ground reaction forces and full-body joint angles), the Layer-Wise Relevance Propagation (LRP) technique reliably demonstrates which variables at what time windows of the gait cycle are most relevant for the characterisation of gait patterns from a certain individual. By measuring the time-resolved contribution of each input variable to the prediction of ML techniques such as DNNs, our method describes the first general framework that enables to understand and interpret non-linear ML methods in (biomechanical) gait analysis and thereby supplies a powerful tool for analysis, diagnosis and treatment of human gait.</jats:p>
- Autoren
- Fabian Horst
- Sebastian Lapuschkin
- Wojciech Samek
- Klaus-Robert Müller
- Wolfgang I Schöllhorn
- DOI
- 10.1038/s41598-019-38748-8
- eISSN
- 2045-2322
- Ausgabe der Veröffentlichung
- 1
- Zeitschrift
- Scientific Reports
- Sprache
- en
- Artikelnummer
- 2391
- Online publication date
- 2019
- Status
- Published online
- Herausgeber
- Springer Science and Business Media LLC
- Herausgeber URL
- http://dx.doi.org/10.1038/s41598-019-38748-8
- Datum der Datenerfassung
- 2022
- Titel
- Explaining the unique nature of individual gait patterns with deep learning
- Ausgabe der Zeitschrift
- 9
Datenquelle: Crossref
- Abstract
- Machine learning (ML) techniques such as (deep) artificial neural networks (DNN) are solving very successfully a plethora of tasks and provide new predictive models for complex physical, chemical, biological and social systems. However, in most cases this comes with the disadvantage of acting as a black box, rarely providing information about what made them arrive at a particular prediction. This black box aspect of ML techniques can be problematic especially in medical diagnoses, so far hampering a clinical acceptance. The present paper studies the uniqueness of individual gait patterns in clinical biomechanics using DNNs. By attributing portions of the model predictions back to the input variables (ground reaction forces and full-body joint angles), the Layer-Wise Relevance Propagation (LRP) technique reliably demonstrates which variables at what time windows of the gait cycle are most relevant for the characterisation of gait patterns from a certain individual. By measuring the time-resolved contribution of each input variable to the prediction of ML techniques such as DNNs, our method describes the first general framework that enables to understand and interpret non-linear ML methods in (biomechanical) gait analysis and thereby supplies a powerful tool for analysis, diagnosis and treatment of human gait.
- Addresses
- Department of Training and Movement Science, Institute of Sport Science, Johannes Gutenberg-University Mainz, Mainz, Rhineland-Palatinate, Germany.
- Autoren
- Fabian Horst
- Sebastian Lapuschkin
- Wojciech Samek
- Klaus-Robert Müller
- Wolfgang I Schöllhorn
- DOI
- 10.1038/s41598-019-38748-8
- eISSN
- 2045-2322
- Externe Identifier
- PubMed Identifier: 30787319
- PubMed Central ID: PMC6382912
- Open access
- true
- ISSN
- 2045-2322
- Ausgabe der Veröffentlichung
- 1
- Zeitschrift
- Scientific reports
- Schlüsselwörter
- Humans
- Gait
- Adult
- Female
- Male
- Young Adult
- Healthy Volunteers
- Biomechanical Phenomena
- Deep Learning
- Sprache
- eng
- Medium
- Electronic
- Online publication date
- 2019
- Open access status
- Open Access
- Paginierung
- 2391
- Datum der Veröffentlichung
- 2019
- Status
- Published
- Publisher licence
- CC BY
- Datum der Datenerfassung
- 2019
- Titel
- Explaining the unique nature of individual gait patterns with deep learning.
- Sub types
- Research Support, Non-U.S. Gov't
- research-article
- Journal Article
- Ausgabe der Zeitschrift
- 9
Files
https://www.nature.com/articles/s41598-019-38748-8.pdf https://europepmc.org/articles/PMC6382912?pdf=render
Datenquelle: Europe PubMed Central
- Abstract
- Machine learning (ML) techniques such as (deep) artificial neural networks (DNN) are solving very successfully a plethora of tasks and provide new predictive models for complex physical, chemical, biological and social systems. However, in most cases this comes with the disadvantage of acting as a black box, rarely providing information about what made them arrive at a particular prediction. This black box aspect of ML techniques can be problematic especially in medical diagnoses, so far hampering a clinical acceptance. The present paper studies the uniqueness of individual gait patterns in clinical biomechanics using DNNs. By attributing portions of the model predictions back to the input variables (ground reaction forces and full-body joint angles), the Layer-Wise Relevance Propagation (LRP) technique reliably demonstrates which variables at what time windows of the gait cycle are most relevant for the characterisation of gait patterns from a certain individual. By measuring the time-resolved contribution of each input variable to the prediction of ML techniques such as DNNs, our method describes the first general framework that enables to understand and interpret non-linear ML methods in (biomechanical) gait analysis and thereby supplies a powerful tool for analysis, diagnosis and treatment of human gait.
- Date of acceptance
- 2019
- Autoren
- Fabian Horst
- Sebastian Lapuschkin
- Wojciech Samek
- Klaus-Robert Müller
- Wolfgang I Schöllhorn
- Autoren-URL
- https://www.ncbi.nlm.nih.gov/pubmed/30787319
- DOI
- 10.1038/s41598-019-38748-8
- eISSN
- 2045-2322
- Externe Identifier
- PubMed Central ID: PMC6382912
- Ausgabe der Veröffentlichung
- 1
- Zeitschrift
- Sci Rep
- Schlüsselwörter
- Adult
- Biomechanical Phenomena
- Deep Learning
- Female
- Gait
- Healthy Volunteers
- Humans
- Male
- Young Adult
- Sprache
- eng
- Country
- England
- Paginierung
- 2391
- PII
- 10.1038/s41598-019-38748-8
- Datum der Veröffentlichung
- 2019
- Status
- Published online
- Datum, an dem der Datensatz öffentlich gemacht wurde
- 2020
- Titel
- Explaining the unique nature of individual gait patterns with deep learning.
- Sub types
- Journal Article
- Research Support, Non-U.S. Gov't
- Ausgabe der Zeitschrift
- 9
Datenquelle: PubMed
- Author's licence
- CC-BY
- Autoren
- Fabian Horst
- Sebastian Lapuschkin
- Wojciech Samek
- Klaus-Robert Müller
- Wolfgang I Schöllhorn
- Hosting institution
- Universitätsbibliothek Mainz
- Sammlungen
- JGU-Publikationen
- Resource version
- Published version
- URN
- urn:nbn:de:hebis:77-publ-589723
- DOI
- 10.1038/s41598-019-38748-8
- Funding acknowledgements
- DFG, Open Access-Publizieren Universität Mainz / Universitätsmedizin
- File(s) embargoed
- false
- Open access
- true
- ISSN
- 2045-2322
- Zeitschrift
- Scientific reports
- Schlüsselwörter
- 796 Sport
- 796 Athletic and outdoor sports and games
- Sprache
- eng
- Open access status
- Open Access
- Paginierung
- Art. 2391
- Datum der Veröffentlichung
- 2019
- Public URL
- https://openscience.ub.uni-mainz.de/handle/20.500.12030/484
- Herausgeber
- Macmillan Publishers Limited, part of Springer Nature
- Herausgeber URL
- http://dx.doi.org/10.1038/s41598-019-38748-8
- Datum der Datenerfassung
- 2019
- Datum, an dem der Datensatz öffentlich gemacht wurde
- 2019
- Zugang
- Public
- Titel
- Explaining the unique nature of individual gait patterns with deep learning
- Ausgabe der Zeitschrift
- 9
Files
horst_fabian-explaining_the-20200928105722868.pdf
Datenquelle: OPENSCIENCE.UB
- Abstract
- Machine learning (ML) techniques such as (deep) artificial neural networks (DNN) are solving very successfully a plethora of tasks and provide new predictive models for complex physical, chemical, biological and social systems. However, in most cases this comes with the disadvantage of acting as a black box, rarely providing information about what made them arrive at a particular prediction. This black box aspect of ML techniques can be problematic especially in medical diagnoses, so far hampering a clinical acceptance. The present paper studies the uniqueness of individual gait patterns in clinical biomechanics using DNNs. By attributing portions of the model predictions back to the input variables (ground reaction forces and full-body joint angles), the Layer-Wise Relevance Propagation (LRP) technique reliably demonstrates which variables at what time windows of the gait cycle are most relevant for the characterisation of gait patterns from a certain individual. By measuring the time-resolved contribution of each input variable to the prediction of ML techniques such as DNNs, our method describes the first general framework that enables to understand and interpret non-linear ML methods in (biomechanical) gait analysis and thereby supplies a powerful tool for analysis, diagnosis and treatment of human gait.
- Autoren
- Fabian Horst
- Sebastian Lapuschkin
- Wojciech Samek
- Klaus-Robert Müller
- Wolfgang I Schöllhorn
- Autoren-URL
- http://arxiv.org/abs/1808.04308v2
- Schlüsselwörter
- cs.LG
- cs.LG
- stat.ML
- Notes
- 17 pages (23 pages including references, 24 pages including references and auxiliary statements, 33 pages including references, auxiliary statements and and supplementary material). 5 figures, 3 tables, 4 supplementary figures, 9 supplementary tables. Accepted for publication at Scientific Reports: https://doi.org/10.1038/s41598-019-38748-8
- Datum der Veröffentlichung
- 2018
- Herausgeber URL
- http://dx.doi.org/10.1038/s41598-019-38748-8
- Datum der Datenerfassung
- 2018
- Datum, an dem der Datensatz öffentlich gemacht wurde
- 2018
- Titel
- Explaining the Unique Nature of Individual Gait Patterns with Deep Learning
Files
1808.04308v2.pdf
Datenquelle: arXiv
- Abstract
- Machine learning (ML) techniques such as (deep) artificial neural networks (DNN) are solving very successfully a plethora of tasks and provide new predictive models for complex physical, chemical, biological and social systems. However, in most cases this comes with the disadvantage of acting as a black box, rarely providing information about what made them arrive at a particular prediction. This black box aspect of ML techniques can be problematic especially in medical diagnoses, so far hampering a clinical acceptance. The present paper studies the uniqueness of individual gait patterns in clinical biomechanics using DNNs. By attributing portions of the model predictions back to the input variables (ground reaction forces and full-body joint angles), the Layer-Wise Relevance Propagation (LRP) technique reliably demonstrates which variables at what time windows of the gait cycle are most relevant for the characterisation of gait patterns from a certain individual. By measuring the time-resolved contribution of each input variable to the prediction of ML techniques such as DNNs, our method describes the first general framework that enables to understand and interpret non-linear ML methods in (biomechanical) gait analysis and thereby supplies a powerful tool for analysis, diagnosis and treatment of human gait.
- Autoren
- Fabian Horst
- Sebastian Lapuschkin
- Wojciech Samek
- Klaus-Robert Müller
- Wolfgang I Schöllhorn
- DOI
- 10.1038/s41598-019-38748-8
- Zeitschrift
- Scientific reports
- Notes
- file: http://www.ncbi.nlm.nih.gov/pubmed/30787319 file: http://www.ncbi.nlm.nih.gov/pubmed/30787319
- Artikelnummer
- 1
- Paginierung
- 2391 - 2391
- Datum der Veröffentlichung
- 2019
- Datum der Datenerfassung
- 2021
- Titel
- Explaining the unique nature of individual gait patterns with deep learning
- Sub types
- article
- Ausgabe der Zeitschrift
- 9
Datenquelle: Manual
- Beziehungen:
- Eigentum von