A review of classification algorithms for EEG-based brain–computer interfaces

By Fabien Lotte, Marco Congedo, Anatole Lécuyer, F. Lamarche, B. Arnaldi. Journal of Neural Engineering, 2007

2,742 citations recorded by the source corpus.

Subfields: Cognitive Neuroscience, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/1a2392702b473b51b80c8511963c25e522dbce1f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Electrical and Electronic Engineering

Citations in the corpus

Learning Representations from EEG with Deep Recurrent-Convolutional Neural Networks (2015, neutral): “State-of-the-art mental state recognition using EEG consists of manual feature selection from continuous time series and applying supervised learning algorithms to learn the discriminative manifold between the states (Lotte & Congedo, 2007; Sub-asi & Ismail Gursoy, 2010).”
A Deep Learning Scheme for Motor Imagery Classification based on Restricted Boltzmann Machines (2017, methodology): “Existing motor imagery classification methods can be categorized into five types [8]: 1) linear classifiers with linear discriminant analysis (LDA) and support vector machine (SVM) being the representative ones; 2) nonlinear Bayesian classifiers such as Bayes quadratic and Hidden Markov Model (HMM);”
EEGNet: a compact convolutional neural network for EEG-based brain–computer interfaces (2016, methodology): “methods [9], a process which often requires significant subject-”
Affect Detection: An Interdisciplinary Review of Models, Methods, and Their Applications (2010, background): “Signal processing [126] and classification algorithms [127] for EEG have been developed in the context of building Brain Computer Interfaces (BCIs), and researchers are constantly seeking ways for developing new approaches to recognizing affective states from EEG and other physiological signals.”
EEG-Based Brain-Computer Interfaces (2017, methodology): “For example, the Lotte et al. [36] review of studies evaluating BCI signal-classiﬁcation algorithms found that most used ofﬂine analyses.”
Silent speech interfaces (2010, background): “Good overviews can be found in (Dornhege et al., 2007) and in (Wolpaw et al., 2002), while Lotte et al. (2007) provides a review of classification algorithms.”
Short-term emotion assessment in a recall paradigm (2009, background): “Current BCI paradigms (Kronegg et al., 2007; Lotte et al., 2007; Vaughan et al., 2003) aim to detect brain activity that corresponds to complex tasks (mental calculus, imagination of finger taping, etc.) not related to the objective of the user (moving a mouse cursor, choosing a letter, etc.).”
Riemannian Approaches in Brain-Computer Interfaces: A Review (2017, neutral): “vector machines (SVM) [22], since they use vectors as input.”
Brain-Computer Interfaces Using Sensorimotor Rhythms: Current State and Future Perspectives (2014, neutral): “various pattern classification techniques [88], [121].”
Decomposing Spatiotemporal Brain Patterns into Topographic Latent Sources (2014, neutral): “Dimensionality reduction methods have a long history in EEG analysis (Lotte et al., 2007; Makeig et al., 2004).”
Signal Processing Approaches to Minimize or Suppress Calibration Time in Oscillatory Activity-Based Brain–Computer Interfaces (2015, methodology): “The LDA classifier uses a linear hyperplane to separate feature vectors from two classes [10].”
Federated Transfer Learning for EEG Signal Classification (2020, methodology): “RELATED WORK 1) Covariance Methods in BCI: When we process and classify EEG signals, one cutting edge technology is to directly estimate and manipulate covariance matrices of EEG signal for source extraction instead of subspaces methods, such as principal component analysis (PCA), independent compon”
Subject-Independent Brain–Computer Interfaces Based on Deep Convolutional Neural Networks (2019, background): “However, unfortunately, the process of calibration is an inconvenient and time-consuming task, which requires approximately 20–30 min for a reliable decoder to be built [24], [25].”
A review of classification algorithms for EEG-based brain–computer interfaces: a 10 year update (2018, background): “We thus advise our readers new to the EEG classification field to start by reading [141], as that paper is more accessible, and the concepts it presented will not be explained again in the current manuscript.”
Flashing characters with famous faces improves ERP-based brain–computer interface performance (2011, background): “Many researchers have addressed this issue by improving (1) the quality of signal processing and the development of novel classiﬁcation techniques (e.g. Krusienski et al (2006), (2008), Blankertz et al (2011); for a review see Lotte et al (2007)) and (2) procedural aspects of the oddball paradigm.”
Regularizing Common Spatial Patterns to Improve BCI Designs: Unified Theory and New Algorithms (2011, methodology): “The log-variances of the spatially filtered EEG signals were then used as input features to an LDA, one of the most efficient classifiers for BCI [2].”