Monte Carlo Dropout for Uncertainty Estimation and Motor Imagery Classification

Motor Imagery (MI)-based Brain–Computer Interfaces (BCIs) have been widely used as an alternative communication channel to patients with severe motor disabilities, achieving high classification accuracy through machine learning techniques. Recently, deep learning techniques have spotlighted the stat...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Daily Milanés-Hermosilla, Rafael Trujillo Codorniú, René López-Baracaldo, Roberto Sagaró-Zamora, Denis Delisle-Rodriguez, John Jairo Villarejo-Mayor, José Ricardo Núñez-Álvarez
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	Brain–Computer Interfaces Monte Carlo dropout motor imagery Shallow Convolutional Neural Network uncertainty estimation Chemical technology TP1-1185
Acceso en línea:	https://doaj.org/article/b665f6d72ba44165a4a48b85e53a98d8
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:b665f6d72ba44165a4a48b85e53a98d8
record_format	dspace
spelling	oai:doaj.org-article:b665f6d72ba44165a4a48b85e53a98d82021-11-11T19:12:54ZMonte Carlo Dropout for Uncertainty Estimation and Motor Imagery Classification10.3390/s212172411424-8220https://doaj.org/article/b665f6d72ba44165a4a48b85e53a98d82021-10-01T00:00:00Zhttps://www.mdpi.com/1424-8220/21/21/7241https://doaj.org/toc/1424-8220Motor Imagery (MI)-based Brain–Computer Interfaces (BCIs) have been widely used as an alternative communication channel to patients with severe motor disabilities, achieving high classification accuracy through machine learning techniques. Recently, deep learning techniques have spotlighted the state-of-the-art of MI-based BCIs. These techniques still lack strategies to quantify predictive uncertainty and may produce overconfident predictions. In this work, methods to enhance the performance of existing MI-based BCIs are proposed in order to obtain a more reliable system for real application scenarios. First, the Monte Carlo dropout (MCD) method is proposed on MI deep neural models to improve classification and provide uncertainty estimation. This approach was implemented using Shallow Convolutional Neural Network (SCNN-MCD) and with an ensemble model (E-SCNN-MCD). As another contribution, to discriminate MI task predictions of high uncertainty, a threshold approach is introduced and tested for both SCNN-MCD and E-SCNN-MCD approaches. The BCI Competition IV Databases 2a and 2b were used to evaluate the proposed methods for both subject-specific and non-subject-specific strategies, obtaining encouraging results for MI recognition.Daily Milanés-HermosillaRafael Trujillo CodorniúRené López-BaracaldoRoberto Sagaró-ZamoraDenis Delisle-RodriguezJohn Jairo Villarejo-MayorJosé Ricardo Núñez-ÁlvarezMDPI AGarticleBrain–Computer InterfacesMonte Carlo dropoutmotor imageryShallow Convolutional Neural Networkuncertainty estimationChemical technologyTP1-1185ENSensors, Vol 21, Iss 7241, p 7241 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Brain–Computer Interfaces Monte Carlo dropout motor imagery Shallow Convolutional Neural Network uncertainty estimation Chemical technology TP1-1185
spellingShingle	Brain–Computer Interfaces Monte Carlo dropout motor imagery Shallow Convolutional Neural Network uncertainty estimation Chemical technology TP1-1185 Daily Milanés-Hermosilla Rafael Trujillo Codorniú René López-Baracaldo Roberto Sagaró-Zamora Denis Delisle-Rodriguez John Jairo Villarejo-Mayor José Ricardo Núñez-Álvarez Monte Carlo Dropout for Uncertainty Estimation and Motor Imagery Classification
description	Motor Imagery (MI)-based Brain–Computer Interfaces (BCIs) have been widely used as an alternative communication channel to patients with severe motor disabilities, achieving high classification accuracy through machine learning techniques. Recently, deep learning techniques have spotlighted the state-of-the-art of MI-based BCIs. These techniques still lack strategies to quantify predictive uncertainty and may produce overconfident predictions. In this work, methods to enhance the performance of existing MI-based BCIs are proposed in order to obtain a more reliable system for real application scenarios. First, the Monte Carlo dropout (MCD) method is proposed on MI deep neural models to improve classification and provide uncertainty estimation. This approach was implemented using Shallow Convolutional Neural Network (SCNN-MCD) and with an ensemble model (E-SCNN-MCD). As another contribution, to discriminate MI task predictions of high uncertainty, a threshold approach is introduced and tested for both SCNN-MCD and E-SCNN-MCD approaches. The BCI Competition IV Databases 2a and 2b were used to evaluate the proposed methods for both subject-specific and non-subject-specific strategies, obtaining encouraging results for MI recognition.
format	article
author	Daily Milanés-Hermosilla Rafael Trujillo Codorniú René López-Baracaldo Roberto Sagaró-Zamora Denis Delisle-Rodriguez John Jairo Villarejo-Mayor José Ricardo Núñez-Álvarez
author_facet	Daily Milanés-Hermosilla Rafael Trujillo Codorniú René López-Baracaldo Roberto Sagaró-Zamora Denis Delisle-Rodriguez John Jairo Villarejo-Mayor José Ricardo Núñez-Álvarez
author_sort	Daily Milanés-Hermosilla
title	Monte Carlo Dropout for Uncertainty Estimation and Motor Imagery Classification
title_short	Monte Carlo Dropout for Uncertainty Estimation and Motor Imagery Classification
title_full	Monte Carlo Dropout for Uncertainty Estimation and Motor Imagery Classification
title_fullStr	Monte Carlo Dropout for Uncertainty Estimation and Motor Imagery Classification
title_full_unstemmed	Monte Carlo Dropout for Uncertainty Estimation and Motor Imagery Classification
title_sort	monte carlo dropout for uncertainty estimation and motor imagery classification
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/b665f6d72ba44165a4a48b85e53a98d8
work_keys_str_mv	AT dailymilaneshermosilla montecarlodropoutforuncertaintyestimationandmotorimageryclassification AT rafaeltrujillocodorniu montecarlodropoutforuncertaintyestimationandmotorimageryclassification AT renelopezbaracaldo montecarlodropoutforuncertaintyestimationandmotorimageryclassification AT robertosagarozamora montecarlodropoutforuncertaintyestimationandmotorimageryclassification AT denisdelislerodriguez montecarlodropoutforuncertaintyestimationandmotorimageryclassification AT johnjairovillarejomayor montecarlodropoutforuncertaintyestimationandmotorimageryclassification AT josericardonunezalvarez montecarlodropoutforuncertaintyestimationandmotorimageryclassification
_version_	1718431606094430208

Monte Carlo Dropout for Uncertainty Estimation and Motor Imagery Classification

Ejemplares similares