Data Augmentation Effects on Highly Imbalanced EEG Datasets for Automatic Detection of Photoparoxysmal Responses
Photosensitivity is a neurological disorder in which a person’s brain produces epileptic discharges, known as Photoparoxysmal Responses (PPRs), when it receives certain visual stimuli. The current standardized diagnosis process used in hospitals consists of submitting the subject to the Intermittent...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-02-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/23/4/2312 |
_version_ | 1797618236043296768 |
---|---|
author | Fernando Moncada Martins Víctor Manuel González Suárez José Ramón Villar Flecha Beatriz García López |
author_facet | Fernando Moncada Martins Víctor Manuel González Suárez José Ramón Villar Flecha Beatriz García López |
author_sort | Fernando Moncada Martins |
collection | DOAJ |
description | Photosensitivity is a neurological disorder in which a person’s brain produces epileptic discharges, known as Photoparoxysmal Responses (PPRs), when it receives certain visual stimuli. The current standardized diagnosis process used in hospitals consists of submitting the subject to the Intermittent Photic Stimulation process and attempting to trigger these phenomena. The brain activity is measured by an Electroencephalogram (EEG), and the clinical specialists manually look for the PPRs that were provoked during the session. Due to the nature of this disorder, long EEG recordings may contain very few PPR segments, meaning that a highly imbalanced dataset is available. To tackle this problem, this research focused on applying Data Augmentation (DA) to create synthetic PPR segments from the real ones, improving the balance of the dataset and, thus, the global performance of the Machine Learning techniques applied for automatic PPR detection. K-Nearest Neighbors and a One-Hidden-Dense-Layer Neural Network were employed to evaluate the performance of this DA stage. The results showed that DA is able to improve the models, making them more robust and more able to generalize. A comparison with the results obtained from a previous experiment also showed a performance improvement of around 20% for the Accuracy and Specificity measurements without Sensitivity suffering any losses. This project is currently being carried out with subjects at Burgos University Hospital, Spain. |
first_indexed | 2024-03-11T08:10:13Z |
format | Article |
id | doaj.art-c46c00eb97f642f8a31d389b51356223 |
institution | Directory Open Access Journal |
issn | 1424-8220 |
language | English |
last_indexed | 2024-03-11T08:10:13Z |
publishDate | 2023-02-01 |
publisher | MDPI AG |
record_format | Article |
series | Sensors |
spelling | doaj.art-c46c00eb97f642f8a31d389b513562232023-11-16T23:13:09ZengMDPI AGSensors1424-82202023-02-01234231210.3390/s23042312Data Augmentation Effects on Highly Imbalanced EEG Datasets for Automatic Detection of Photoparoxysmal ResponsesFernando Moncada Martins0Víctor Manuel González Suárez1José Ramón Villar Flecha2Beatriz García López3Electrical Engineering Department, University of Oviedo, 33203 Gijón, SpainElectrical Engineering Department, University of Oviedo, 33203 Gijón, SpainComputer Science Department, University of Oviedo, 33003 Oviedo, SpainNeurophysiology Department, University Hospital of Burgos, 09006 Burgos, SpainPhotosensitivity is a neurological disorder in which a person’s brain produces epileptic discharges, known as Photoparoxysmal Responses (PPRs), when it receives certain visual stimuli. The current standardized diagnosis process used in hospitals consists of submitting the subject to the Intermittent Photic Stimulation process and attempting to trigger these phenomena. The brain activity is measured by an Electroencephalogram (EEG), and the clinical specialists manually look for the PPRs that were provoked during the session. Due to the nature of this disorder, long EEG recordings may contain very few PPR segments, meaning that a highly imbalanced dataset is available. To tackle this problem, this research focused on applying Data Augmentation (DA) to create synthetic PPR segments from the real ones, improving the balance of the dataset and, thus, the global performance of the Machine Learning techniques applied for automatic PPR detection. K-Nearest Neighbors and a One-Hidden-Dense-Layer Neural Network were employed to evaluate the performance of this DA stage. The results showed that DA is able to improve the models, making them more robust and more able to generalize. A comparison with the results obtained from a previous experiment also showed a performance improvement of around 20% for the Accuracy and Specificity measurements without Sensitivity suffering any losses. This project is currently being carried out with subjects at Burgos University Hospital, Spain.https://www.mdpi.com/1424-8220/23/4/2312electroencephalographyEEGPhotoparoxysmal ResponsePPRMachine LearningData Augmentation |
spellingShingle | Fernando Moncada Martins Víctor Manuel González Suárez José Ramón Villar Flecha Beatriz García López Data Augmentation Effects on Highly Imbalanced EEG Datasets for Automatic Detection of Photoparoxysmal Responses Sensors electroencephalography EEG Photoparoxysmal Response PPR Machine Learning Data Augmentation |
title | Data Augmentation Effects on Highly Imbalanced EEG Datasets for Automatic Detection of Photoparoxysmal Responses |
title_full | Data Augmentation Effects on Highly Imbalanced EEG Datasets for Automatic Detection of Photoparoxysmal Responses |
title_fullStr | Data Augmentation Effects on Highly Imbalanced EEG Datasets for Automatic Detection of Photoparoxysmal Responses |
title_full_unstemmed | Data Augmentation Effects on Highly Imbalanced EEG Datasets for Automatic Detection of Photoparoxysmal Responses |
title_short | Data Augmentation Effects on Highly Imbalanced EEG Datasets for Automatic Detection of Photoparoxysmal Responses |
title_sort | data augmentation effects on highly imbalanced eeg datasets for automatic detection of photoparoxysmal responses |
topic | electroencephalography EEG Photoparoxysmal Response PPR Machine Learning Data Augmentation |
url | https://www.mdpi.com/1424-8220/23/4/2312 |
work_keys_str_mv | AT fernandomoncadamartins dataaugmentationeffectsonhighlyimbalancedeegdatasetsforautomaticdetectionofphotoparoxysmalresponses AT victormanuelgonzalezsuarez dataaugmentationeffectsonhighlyimbalancedeegdatasetsforautomaticdetectionofphotoparoxysmalresponses AT joseramonvillarflecha dataaugmentationeffectsonhighlyimbalancedeegdatasetsforautomaticdetectionofphotoparoxysmalresponses AT beatrizgarcialopez dataaugmentationeffectsonhighlyimbalancedeegdatasetsforautomaticdetectionofphotoparoxysmalresponses |