Wearable Regionally Trained AI-Enabled Bruxism-Detection System
Sleep Bruxism (SB) and Awake Bruxism (AB) can cause severe discomfort, exhaustion, and problems with day-to-day functioning, including poor sleep and bad performance at work. This emphasizes the significance of early identification and treatment of bruxism. To date, some tools like mouthpieces have...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10848082/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832584003671031808 |
|---|---|
| author | Anusha Ishtiaq Jahanzeb Gul Zia Mohy Ud Din Azhar Imran Khalil El Hindi |
| author_facet | Anusha Ishtiaq Jahanzeb Gul Zia Mohy Ud Din Azhar Imran Khalil El Hindi |
| author_sort | Anusha Ishtiaq |
| collection | DOAJ |
| description | Sleep Bruxism (SB) and Awake Bruxism (AB) can cause severe discomfort, exhaustion, and problems with day-to-day functioning, including poor sleep and bad performance at work. This emphasizes the significance of early identification and treatment of bruxism. To date, some tools like mouthpieces have been designed for teeth protection. However, they are not user-friendly due to their internal placement in the mouth. Bruxers require a gadget that not only identifies and continually monitors their bruxism activity, but also alerts them. In this study, a wearable EMG-based device has been designed to monitor and detect jaw clenching in the supine position using EMG of the two facial muscles, Temporalis and Masseter. This study purposely found which muscle varies most with bruxism activity. The EMG signals’ data of 30 regional subjects, with 5 trials each, have been acquired and pre-processed using filters and three data oversampling techniques, SMOTE, SMOTE-ENN, and ADSYN. The augmented data has been trained, validated, and tested on six machine-learning classifiers and three deep-learning models. The Recurrent Neural Network provided the highest accuracy 0.99 and a recall value 0.98 for the temporalis muscle dataset. The other eight classifiers have provided accuracies in descending order such as Convolutional Neural Network, Long Short-Term Memory, k-Nearest Neighbors, and Decision Tree 0.98; Logistic Regression 0.96; Support Vector Machine 0.97, and Naïve Bayes 0.89, respectively. The module has been tested on several participants, and bruxism is identified when they do jaw clenching or teeth grinding. In the future, the size of the gadget could be miniaturized to ensure the users’ comfort level. |
| format | Article |
| id | doaj-art-e04f9ff66ea94939ba64a1d5a3b2868f |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-e04f9ff66ea94939ba64a1d5a3b2868f2025-01-28T00:01:16ZengIEEEIEEE Access2169-35362025-01-0113155031552810.1109/ACCESS.2025.353236010848082Wearable Regionally Trained AI-Enabled Bruxism-Detection SystemAnusha Ishtiaq0Jahanzeb Gul1https://orcid.org/0000-0002-1230-2546Zia Mohy Ud Din2https://orcid.org/0000-0001-5756-7284Azhar Imran3https://orcid.org/0000-0003-3598-2780Khalil El Hindi4https://orcid.org/0000-0003-2457-9961Department of Biomedical Engineering, Air University, Islamabad, PakistanDepartment of Electronic Engineering, Maynooth University, Maynooth, IrelandDepartment of Biomedical Engineering, Air University, Islamabad, PakistanDepartment of Creative Technologies, Faculty of Computing and AI, Air University, Islamabad, PakistanDepartment of Computer Science, College of Computer and Information Sciences, King Saud University, Riyadh, Saudi ArabiaSleep Bruxism (SB) and Awake Bruxism (AB) can cause severe discomfort, exhaustion, and problems with day-to-day functioning, including poor sleep and bad performance at work. This emphasizes the significance of early identification and treatment of bruxism. To date, some tools like mouthpieces have been designed for teeth protection. However, they are not user-friendly due to their internal placement in the mouth. Bruxers require a gadget that not only identifies and continually monitors their bruxism activity, but also alerts them. In this study, a wearable EMG-based device has been designed to monitor and detect jaw clenching in the supine position using EMG of the two facial muscles, Temporalis and Masseter. This study purposely found which muscle varies most with bruxism activity. The EMG signals’ data of 30 regional subjects, with 5 trials each, have been acquired and pre-processed using filters and three data oversampling techniques, SMOTE, SMOTE-ENN, and ADSYN. The augmented data has been trained, validated, and tested on six machine-learning classifiers and three deep-learning models. The Recurrent Neural Network provided the highest accuracy 0.99 and a recall value 0.98 for the temporalis muscle dataset. The other eight classifiers have provided accuracies in descending order such as Convolutional Neural Network, Long Short-Term Memory, k-Nearest Neighbors, and Decision Tree 0.98; Logistic Regression 0.96; Support Vector Machine 0.97, and Naïve Bayes 0.89, respectively. The module has been tested on several participants, and bruxism is identified when they do jaw clenching or teeth grinding. In the future, the size of the gadget could be miniaturized to ensure the users’ comfort level.https://ieeexplore.ieee.org/document/10848082/Bruxismclassificationdeep learningmasseter muscletemporalis muscle |
| spellingShingle | Anusha Ishtiaq Jahanzeb Gul Zia Mohy Ud Din Azhar Imran Khalil El Hindi Wearable Regionally Trained AI-Enabled Bruxism-Detection System IEEE Access Bruxism classification deep learning masseter muscle temporalis muscle |
| title | Wearable Regionally Trained AI-Enabled Bruxism-Detection System |
| title_full | Wearable Regionally Trained AI-Enabled Bruxism-Detection System |
| title_fullStr | Wearable Regionally Trained AI-Enabled Bruxism-Detection System |
| title_full_unstemmed | Wearable Regionally Trained AI-Enabled Bruxism-Detection System |
| title_short | Wearable Regionally Trained AI-Enabled Bruxism-Detection System |
| title_sort | wearable regionally trained ai enabled bruxism detection system |
| topic | Bruxism classification deep learning masseter muscle temporalis muscle |
| url | https://ieeexplore.ieee.org/document/10848082/ |
| work_keys_str_mv | AT anushaishtiaq wearableregionallytrainedaienabledbruxismdetectionsystem AT jahanzebgul wearableregionallytrainedaienabledbruxismdetectionsystem AT ziamohyuddin wearableregionallytrainedaienabledbruxismdetectionsystem AT azharimran wearableregionallytrainedaienabledbruxismdetectionsystem AT khalilelhindi wearableregionallytrainedaienabledbruxismdetectionsystem |