Explainable Artificial Intelligence (XAI) in Pain Research: Understanding the Role of Electrodermal Activity for Automated Pain Recognition
Artificial intelligence and especially deep learning methods have achieved outstanding results for various applications in the past few years. Pain recognition is one of them, as various models have been proposed to replace the previous gold standard with an automated and objective assessment. While...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-02-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/4/1959 |
| _version_ | 1827755546455834624 |
|---|---|
| author | Philip Gouverneur Frédéric Li Kimiaki Shirahama Luisa Luebke Wacław M. Adamczyk Tibor M. Szikszay Kerstin Luedtke Marcin Grzegorzek |
| author_facet | Philip Gouverneur Frédéric Li Kimiaki Shirahama Luisa Luebke Wacław M. Adamczyk Tibor M. Szikszay Kerstin Luedtke Marcin Grzegorzek |
| author_sort | Philip Gouverneur |
| collection | DOAJ |
| description | Artificial intelligence and especially deep learning methods have achieved outstanding results for various applications in the past few years. Pain recognition is one of them, as various models have been proposed to replace the previous gold standard with an automated and objective assessment. While the accuracy of such models could be increased incrementally, the understandability and transparency of these systems have not been the main focus of the research community thus far. Thus, in this work, several outcomes and insights of explainable artificial intelligence applied to the electrodermal activity sensor data of the PainMonit and BioVid Heat Pain Database are presented. For this purpose, the importance of hand-crafted features is evaluated using recursive feature elimination based on impurity scores in Random Forest (RF) models. Additionally, Gradient-weighted class activation mapping is applied to highlight the most impactful features learned by deep learning models. Our studies highlight the following insights: (1) Very simple hand-crafted features can yield comparative performances to deep learning models for pain recognition, especially when properly selected with recursive feature elimination. Thus, the use of complex neural networks should be questioned in pain recognition, especially considering their computational costs; and (2) both traditional feature engineering and deep feature learning approaches rely on simple characteristics of the input time-series data to make their decision in the context of automated pain recognition. |
| first_indexed | 2024-03-11T08:10:52Z |
| format | Article |
| id | doaj.art-70ed2c6bdff34840a9afe8214b346aa8 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-11T08:10:52Z |
| publishDate | 2023-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-70ed2c6bdff34840a9afe8214b346aa82023-11-16T23:08:26ZengMDPI AGSensors1424-82202023-02-01234195910.3390/s23041959Explainable Artificial Intelligence (XAI) in Pain Research: Understanding the Role of Electrodermal Activity for Automated Pain RecognitionPhilip Gouverneur0Frédéric Li1Kimiaki Shirahama2Luisa Luebke3Wacław M. Adamczyk4Tibor M. Szikszay5Kerstin Luedtke6Marcin Grzegorzek7Institute of Medical Informatics, University of Lübeck, Ratzeburger Allee 160, 23562 Lübeck, GermanyInstitute of Medical Informatics, University of Lübeck, Ratzeburger Allee 160, 23562 Lübeck, GermanyFaculty of Informatics, Kindai University, Higashiosaka 577-8502, Osaka, JapanDepartment of Physiotherapy, Pain and Exercise Research Luebeck (P.E.R.L.), Institute of Health Sciences, University of Lübeck, Ratzeburger Allee 160, 23562 Lübeck, GermanyDepartment of Physiotherapy, Pain and Exercise Research Luebeck (P.E.R.L.), Institute of Health Sciences, University of Lübeck, Ratzeburger Allee 160, 23562 Lübeck, GermanyDepartment of Physiotherapy, Pain and Exercise Research Luebeck (P.E.R.L.), Institute of Health Sciences, University of Lübeck, Ratzeburger Allee 160, 23562 Lübeck, GermanyDepartment of Physiotherapy, Pain and Exercise Research Luebeck (P.E.R.L.), Institute of Health Sciences, University of Lübeck, Ratzeburger Allee 160, 23562 Lübeck, GermanyInstitute of Medical Informatics, University of Lübeck, Ratzeburger Allee 160, 23562 Lübeck, GermanyArtificial intelligence and especially deep learning methods have achieved outstanding results for various applications in the past few years. Pain recognition is one of them, as various models have been proposed to replace the previous gold standard with an automated and objective assessment. While the accuracy of such models could be increased incrementally, the understandability and transparency of these systems have not been the main focus of the research community thus far. Thus, in this work, several outcomes and insights of explainable artificial intelligence applied to the electrodermal activity sensor data of the PainMonit and BioVid Heat Pain Database are presented. For this purpose, the importance of hand-crafted features is evaluated using recursive feature elimination based on impurity scores in Random Forest (RF) models. Additionally, Gradient-weighted class activation mapping is applied to highlight the most impactful features learned by deep learning models. Our studies highlight the following insights: (1) Very simple hand-crafted features can yield comparative performances to deep learning models for pain recognition, especially when properly selected with recursive feature elimination. Thus, the use of complex neural networks should be questioned in pain recognition, especially considering their computational costs; and (2) both traditional feature engineering and deep feature learning approaches rely on simple characteristics of the input time-series data to make their decision in the context of automated pain recognition.https://www.mdpi.com/1424-8220/23/4/1959pain recognitionmachine learningdeep learninghand-crafted featuresphysiological signalspain perception |
| spellingShingle | Philip Gouverneur Frédéric Li Kimiaki Shirahama Luisa Luebke Wacław M. Adamczyk Tibor M. Szikszay Kerstin Luedtke Marcin Grzegorzek Explainable Artificial Intelligence (XAI) in Pain Research: Understanding the Role of Electrodermal Activity for Automated Pain Recognition Sensors pain recognition machine learning deep learning hand-crafted features physiological signals pain perception |
| title | Explainable Artificial Intelligence (XAI) in Pain Research: Understanding the Role of Electrodermal Activity for Automated Pain Recognition |
| title_full | Explainable Artificial Intelligence (XAI) in Pain Research: Understanding the Role of Electrodermal Activity for Automated Pain Recognition |
| title_fullStr | Explainable Artificial Intelligence (XAI) in Pain Research: Understanding the Role of Electrodermal Activity for Automated Pain Recognition |
| title_full_unstemmed | Explainable Artificial Intelligence (XAI) in Pain Research: Understanding the Role of Electrodermal Activity for Automated Pain Recognition |
| title_short | Explainable Artificial Intelligence (XAI) in Pain Research: Understanding the Role of Electrodermal Activity for Automated Pain Recognition |
| title_sort | explainable artificial intelligence xai in pain research understanding the role of electrodermal activity for automated pain recognition |
| topic | pain recognition machine learning deep learning hand-crafted features physiological signals pain perception |
| url | https://www.mdpi.com/1424-8220/23/4/1959 |
| work_keys_str_mv | AT philipgouverneur explainableartificialintelligencexaiinpainresearchunderstandingtheroleofelectrodermalactivityforautomatedpainrecognition AT fredericli explainableartificialintelligencexaiinpainresearchunderstandingtheroleofelectrodermalactivityforautomatedpainrecognition AT kimiakishirahama explainableartificialintelligencexaiinpainresearchunderstandingtheroleofelectrodermalactivityforautomatedpainrecognition AT luisaluebke explainableartificialintelligencexaiinpainresearchunderstandingtheroleofelectrodermalactivityforautomatedpainrecognition AT wacławmadamczyk explainableartificialintelligencexaiinpainresearchunderstandingtheroleofelectrodermalactivityforautomatedpainrecognition AT tibormszikszay explainableartificialintelligencexaiinpainresearchunderstandingtheroleofelectrodermalactivityforautomatedpainrecognition AT kerstinluedtke explainableartificialintelligencexaiinpainresearchunderstandingtheroleofelectrodermalactivityforautomatedpainrecognition AT marcingrzegorzek explainableartificialintelligencexaiinpainresearchunderstandingtheroleofelectrodermalactivityforautomatedpainrecognition |