Dataset Information

Deep learning for EEG-based Motor Imagery classification: Accuracy-cost trade-off.

ABSTRACT: Electroencephalography (EEG) datasets are often small and high dimensional, owing to cumbersome recording processes. In these conditions, powerful machine learning techniques are essential to deal with the large amount of information and overcome the curse of dimensionality. Artificial Neural Networks (ANNs) have achieved promising performance in EEG-based Brain-Computer Interface (BCI) applications, but they involve computationally intensive training algorithms and hyperparameter optimization methods. Thus, an awareness of the quality-cost trade-off, although usually overlooked, is highly beneficial. In this paper, we apply a hyperparameter optimization procedure based on Genetic Algorithms to Convolutional Neural Networks (CNNs), Feed-Forward Neural Networks (FFNNs), and Recurrent Neural Networks (RNNs), all of them purposely shallow. We compare their relative quality and energy-time cost, but we also analyze the variability in the structural complexity of networks of the same type with similar accuracies. The experimental results show that the optimization procedure improves accuracy in all models, and that CNN models with only one hidden convolutional layer can equal or slightly outperform a 6-layer Deep Belief Network. FFNN and RNN were not able to reach the same quality, although the cost was significantly lower. The results also highlight the fact that size within the same type of network is not necessarily correlated with accuracy, as smaller models can and do match, or even surpass, bigger ones in performance. In this regard, overfitting is likely a contributing factor since deep learning approaches struggle with limited training examples.

SUBMITTER: Leon J

PROVIDER: S-EPMC7289369 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Deep learning for EEG-based Motor Imagery classification: Accuracy-cost trade-off.

León Javier J Escobar Juan José JJ Ortiz Andrés A Ortega Julio J González Jesús J Martín-Smith Pedro P Gan John Q JQ Damas Miguel M

PloS one 20200611 6

Electroencephalography (EEG) datasets are often small and high dimensional, owing to cumbersome recording processes. In these conditions, powerful machine learning techniques are essential to deal with the large amount of information and overcome the curse of dimensionality. Artificial Neural Networks (ANNs) have achieved promising performance in EEG-based Brain-Computer Interface (BCI) applications, but they involve computationally intensive training algorithms and hyperparameter optimization m ...[more]

PMID: 32525885

Similar Datasets

Project description:Motor Imagery Brain-Computer Interfaces (MI-BCIs) are AI-driven systems that capture brain activity patterns associated with mental imagination of movement and convert them into commands for external devices. Traditionally, MI-BCIs operate on Machine Learning (ML) algorithms, which require extensive signal processing and feature engineering to extract changes in sensorimotor rhythms (SMR). In recent years, Deep Learning (DL) models have gained popularity for EEG classification as they provide a solution for automatic extraction of spatio-temporal features in the signals. However, past BCI studies that employed DL models, only attempted them with a small group of participants, without investigating the effectiveness of this approach for different user groups such as inefficient users. BCI inefficiency is a known and unsolved problem within BCI literature, generally defined as the inability of the user to produce the desired SMR patterns for the BCI classifier. In this study, we evaluated the effectiveness of DL models in capturing MI features particularly in the inefficient users. EEG signals from 54 subjects who performed a MI task of left- or right-hand grasp were recorded to compare the performance of two classification approaches; a ML approach vs. a DL approach. In the ML approach, Common Spatial Patterns (CSP) was used for feature extraction and then Linear Discriminant Analysis (LDA) model was employed for binary classification of the MI task. In the DL approach, a Convolutional Neural Network (CNN) model was constructed on the raw EEG signals. Additionally, subjects were divided into high vs. low performers based on their online BCI accuracy and the difference between the two classifiers' performance was compared between groups. Our results showed that the CNN model improved the classification accuracy for all subjects within the range of 2.37 to 28.28%, but more importantly, this improvement was significantly larger for low performers. Our findings show promise for employment of DL models on raw EEG signals in future MI-BCI systems, particularly for BCI inefficient users who are unable to produce desired sensorimotor patterns for conventional ML approaches.

Project description:Motor imagery is based on the volitional modulation of sensorimotor rhythms (SMRs); however, the sensorimotor processes in patients with amyotrophic lateral sclerosis (ALS) are impaired, leading to degenerated motor imagery ability. Thus, motor imagery classification in ALS patients has been considered challenging in the brain-computer interface (BCI) community. In this study, we address this critical issue by introducing the Grassberger-Procaccia and Higuchi's methods to estimate the fractal dimensions (GPFD and HFD, respectively) of the electroencephalography (EEG) signals from ALS patients. Moreover, a Fisher's criterion-based channel selection strategy is proposed to automatically determine the best patient-dependent channel configuration from 30 EEG recording sites. An EEG data collection paradigm is designed to collect the EEG signal of resting state and the imagination of three movements, including right hand grasping (RH), left hand grasping (LH), and left foot stepping (LF). Five late-stage ALS patients without receiving any SMR training participated in this study. Experimental results show that the proposed GPFD feature is not only superior to the previously-used SMR features (mu and beta band powers of EEG from sensorimotor cortex) but also better than HFD. The accuracies achieved by the SMR features are not satisfactory (all lower than 80%) in all binary classification tasks, including RH imagery vs. resting, LH imagery vs. resting, and LF imagery vs. resting. For the discrimination between RH imagery and resting, the average accuracies of GPFD in 30-channel (without channel selection) and top-five-channel configurations are 95.25% and 93.50%, respectively. When using only one channel (the best channel among the 30), a high accuracy of 91.00% can still be achieved by the GPFD feature and a linear discriminant analysis (LDA) classifier. The results also demonstrate that the proposed Fisher's criterion-based channel selection is capable of removing a large amount of redundant and noisy EEG channels. The proposed GPFD feature extraction combined with the channel selection strategy can be used as the basis for further developing high-accuracy and high-usability motor imagery BCI systems from which the patients with ALS can really benefit.

Project description:Introduction Electroencephalogram (EEG)-based motor imagery (MI) classification is an important aspect in brain-computer interfaces (BCIs), which bridges between neural system and computer devices decoding brain signals into recognizable machine commands. However, due to the small number of training samples of MI electroencephalogram (MI-EEG) for a single subject and the great individual differences of MI-EEG among different subjects, the generalization and accuracy of the model on the specific MI task may be poor. Methods To solve these problems, an adaptive cross-subject transfer learning algorithm is proposed, which is based on kernel mean matching (KMM) and transfer learning adaptive boosting (TrAdaBoost) method. First, the common spatial pattern (CSP) is used to extract the spatial features. Then, in order to make the feature distribution more similar among different subjects, the KMM algorithm is used to compute a sample weight matrix for aligning the mean between source and target domains and reducing distribution differences among different subjects. Finally, the sample weight matrix from KMM is used as the initialization weight of TrAdaBoost, and then TrAdaBoost is used to adaptively select source domain samples that are closer to the target task distribution to assist in building a classification model. Results In order to verify the effectiveness and feasibility of the proposed method, the algorithm is applied to BCI Competition IV datasets and in-house datasets. The results show that the average classification accuracy of the proposed method on the public datasets is 89.1%, and the average classification accuracy on the in-house datasets is 80.4%. Discussion Compared with the existing methods, the proposed method effectively improves the classification accuracy of MI-EEG signals. At the same time, this paper also applies the proposed algorithm to the in-house dataset, the results verify the effectiveness of the algorithm again, and the results of this study have certain clinical guiding significance for brain rehabilitation.

Project description:Simple Summary According to skin disease reports by healthcare organizations, the number of cases of skin disease is growing gradually over the years globally. In skin disease diagnosis, dermatologists examine skin cells by using a dermatoscope. Due to the global shortage of expert dermatologists, mainly in developing countries, an accurate early skin disease diagnosis is not possible. To automate the examination of skin disease images, computer-aided diagnosis-based tools are used in healthcare and medical environments. Computer-aided diagnosis employs machine learning including deep learning models on skin disease images to detect and classify skin diseases. The present work proposes a deep learning-based model to accurately detect skin diseases and classify them into a family of skin diseases using skin disease images. The proposed system demonstrated a performance improvement of 4% accuracy for skin disease detection and 9% accuracy for skin disease classification compared to the existing deep learning-based models. The proposed computer-aided tool can be used as an early skin diagnosis tool to assist dermatologists in healthcare and medical environments. Abstract Deep learning-based models have been employed for the detection and classification of skin diseases through medical imaging. However, deep learning-based models are not effective for rare skin disease detection and classification. This is mainly due to the reason that rare skin disease has very a smaller number of data samples. Thus, the dataset will be highly imbalanced, and due to the bias in learning, most of the models give better performances. The deep learning models are not effective in detecting the affected tiny portions of skin disease in the overall regions of the image. This paper presents an attention-cost-sensitive deep learning-based feature fusion ensemble meta-classifier approach for skin cancer detection and classification. Cost weights are included in the deep learning models to handle the data imbalance during training. To effectively learn the optimal features from the affected tiny portions of skin image samples, attention is integrated into the deep learning models. The features from the finetuned models are extracted and the dimensionality of the features was further reduced by using a kernel-based principal component (KPCA) analysis. The reduced features of the deep learning-based finetuned models are fused and passed into ensemble meta-classifiers for skin disease detection and classification. The ensemble meta-classifier is a two-stage model. The first stage performs the prediction of skin disease and the second stage performs the classification by considering the prediction of the first stage as features. Detailed analysis of the proposed approach is demonstrated for both skin disease detection and skin disease classification. The proposed approach demonstrated an accuracy of 99% on skin disease detection and 99% on skin disease classification. In all the experimental settings, the proposed approach outperformed the existing methods and demonstrated a performance improvement of 4% accuracy for skin disease detection and 9% accuracy for skin disease classification. The proposed approach can be used as a computer-aided diagnosis (CAD) tool for the early diagnosis of skin cancer detection and classification in healthcare and medical environments. The tool can accurately detect skin diseases and classify the skin disease into their skin disease family.

Dataset Information

Deep learning for EEG-based Motor Imagery classification: Accuracy-cost trade-off.

Publications

Deep learning for EEG-based Motor Imagery classification: Accuracy-cost trade-off.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets