High-speed predictions of post-burn contraction using a neural network trained on 2D-finite element simulations
Severe burn injuries often lead to skin contraction, leading to stresses in and around the damaged skin region. If this contraction leads to impaired joint mobility, one speaks of contracture. To optimize treatment, a mathematical model, that is based on finite element methods, is developed. Since t...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2023-01-01
|
| Series: | Frontiers in Applied Mathematics and Statistics |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fams.2023.1098242/full |
| _version_ | 1811175316365770752 |
|---|---|
| author | Ginger Egberts Ginger Egberts Fred Vermolen Paul van Zuijlen Paul van Zuijlen Paul van Zuijlen |
| author_facet | Ginger Egberts Ginger Egberts Fred Vermolen Paul van Zuijlen Paul van Zuijlen Paul van Zuijlen |
| author_sort | Ginger Egberts |
| collection | DOAJ |
| description | Severe burn injuries often lead to skin contraction, leading to stresses in and around the damaged skin region. If this contraction leads to impaired joint mobility, one speaks of contracture. To optimize treatment, a mathematical model, that is based on finite element methods, is developed. Since the finite element-based simulation of skin contraction can be expensive from a computational point of view, we use machine learning to replace these simulations such that we have a cheap alternative. The current study deals with a feed-forward neural network that we trained with 2D finite element simulations based on morphoelasticity. We focus on the evolution of the scar shape, wound area, and total strain energy, a measure of discomfort, over time. The results show average goodness of fit (R2) of 0.9979 and a tremendous speedup of 1815000X. Further, we illustrate the applicability of the neural network in an online medical app that takes the patient's age into account. |
| first_indexed | 2024-04-10T19:34:02Z |
| format | Article |
| id | doaj.art-08cc4dd934d348e7aedf7df781d8788f |
| institution | Directory Open Access Journal |
| issn | 2297-4687 |
| language | English |
| last_indexed | 2024-04-10T19:34:02Z |
| publishDate | 2023-01-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Applied Mathematics and Statistics |
| spelling | doaj.art-08cc4dd934d348e7aedf7df781d8788f2023-01-30T08:27:20ZengFrontiers Media S.A.Frontiers in Applied Mathematics and Statistics2297-46872023-01-01910.3389/fams.2023.10982421098242High-speed predictions of post-burn contraction using a neural network trained on 2D-finite element simulationsGinger Egberts0Ginger Egberts1Fred Vermolen2Paul van Zuijlen3Paul van Zuijlen4Paul van Zuijlen5Delft Institute of Applied Mathematics, Delft University of Technology, Delft, NetherlandsResearch Group Computational Mathematics (CMAT), Department of Mathematics and Statistics, University of Hasselt, Hasselt, BelgiumResearch Group Computational Mathematics (CMAT), Department of Mathematics and Statistics, University of Hasselt, Hasselt, BelgiumBurn Centre and Department of Plastic, Reconstructive and Hand Surgery, Red Cross Hospital, Beverwijk, NetherlandsDepartment of Plastic, Reconstructive and Hand Surgery, Amsterdam UMC, Location VUmc, Amsterdam Movement Sciences, Amsterdam, NetherlandsPediatric Surgical Centre, Emma Children's Hospital, Amsterdam UMC, Location AMC and VUmc, Amsterdam, NetherlandsSevere burn injuries often lead to skin contraction, leading to stresses in and around the damaged skin region. If this contraction leads to impaired joint mobility, one speaks of contracture. To optimize treatment, a mathematical model, that is based on finite element methods, is developed. Since the finite element-based simulation of skin contraction can be expensive from a computational point of view, we use machine learning to replace these simulations such that we have a cheap alternative. The current study deals with a feed-forward neural network that we trained with 2D finite element simulations based on morphoelasticity. We focus on the evolution of the scar shape, wound area, and total strain energy, a measure of discomfort, over time. The results show average goodness of fit (R2) of 0.9979 and a tremendous speedup of 1815000X. Further, we illustrate the applicability of the neural network in an online medical app that takes the patient's age into account.https://www.frontiersin.org/articles/10.3389/fams.2023.1098242/fullmachine learningpost-burn scar contractionmorphoelasticityfeed–forward neural networkonline applicationMonte Carlo simulations |
| spellingShingle | Ginger Egberts Ginger Egberts Fred Vermolen Paul van Zuijlen Paul van Zuijlen Paul van Zuijlen High-speed predictions of post-burn contraction using a neural network trained on 2D-finite element simulations Frontiers in Applied Mathematics and Statistics machine learning post-burn scar contraction morphoelasticity feed–forward neural network online application Monte Carlo simulations |
| title | High-speed predictions of post-burn contraction using a neural network trained on 2D-finite element simulations |
| title_full | High-speed predictions of post-burn contraction using a neural network trained on 2D-finite element simulations |
| title_fullStr | High-speed predictions of post-burn contraction using a neural network trained on 2D-finite element simulations |
| title_full_unstemmed | High-speed predictions of post-burn contraction using a neural network trained on 2D-finite element simulations |
| title_short | High-speed predictions of post-burn contraction using a neural network trained on 2D-finite element simulations |
| title_sort | high speed predictions of post burn contraction using a neural network trained on 2d finite element simulations |
| topic | machine learning post-burn scar contraction morphoelasticity feed–forward neural network online application Monte Carlo simulations |
| url | https://www.frontiersin.org/articles/10.3389/fams.2023.1098242/full |
| work_keys_str_mv | AT gingeregberts highspeedpredictionsofpostburncontractionusinganeuralnetworktrainedon2dfiniteelementsimulations AT gingeregberts highspeedpredictionsofpostburncontractionusinganeuralnetworktrainedon2dfiniteelementsimulations AT fredvermolen highspeedpredictionsofpostburncontractionusinganeuralnetworktrainedon2dfiniteelementsimulations AT paulvanzuijlen highspeedpredictionsofpostburncontractionusinganeuralnetworktrainedon2dfiniteelementsimulations AT paulvanzuijlen highspeedpredictionsofpostburncontractionusinganeuralnetworktrainedon2dfiniteelementsimulations AT paulvanzuijlen highspeedpredictionsofpostburncontractionusinganeuralnetworktrainedon2dfiniteelementsimulations |