Article
Chemistry, Analytical
Xuying Wang, Rui Yang, Mengjie Huang
Summary: This study proposes an unsupervised deep-transfer-learning-based method to address the limitations of brain-computer interface (BCI) systems. By aligning data in Euclidean space and extracting features using common spatial pattern (CSP), the method achieves EEG signal classification through deep convolutional neural network (CNN). Experimental results demonstrate the effectiveness of the proposed method.
Article
Engineering, Multidisciplinary
Pasquale Arpaia, Damien Coyle, Francesco Donnarumma, Antonio Esposito, Angela Natalizio, Marco Parvis
Summary: This paper presents a wearable brain-computer interface that enhances motor imagery training through neurofeedback in extended reality. Various feedback modalities, including visual and vibrotactile, were evaluated either singularly or simultaneously. The results showed statistically significant improvement in performance over multiple sessions, demonstrating the functionality of the motor imagery-based instrument even with minimal equipment. The best feedback modality was found to be subject-dependent, with classification accuracy exceeding 80% in some cases.
Article
Mathematics
Md. Khademul Islam Molla, Sakir Ahamed, Ahmed M. M. Almassri, Hiroaki Wagatsuma
Summary: This paper presents a method for recording electrical activities of the human brain using electroencephalography. It decomposes the raw EEG signals and extracts spatial features to classify motor imagery tasks. The proposed method achieves higher classification accuracy in BCI implementation compared to other algorithms.
Article
Biology
Huiying Li, Dongxue Zhang, Jingmeng Xie
Summary: A novel dual-attention-based adversarial network for motor imagery classification (MI-DABAN) is proposed, which leverages multiple subjects' knowledge to improve a single subject's classification performance. The method employs a clever adversarial learning method and two unshared attention blocks, resulting in effective and superior classification performance.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Article
Chemistry, Analytical
Mateo Tobon-Henao, Andres Alvarez-Meza, German Castellanos-Dominguez
Summary: The EEG-based motor imagery paradigm is widely studied in the field of Brain-Computer Interface (BCI) development, but it faces challenges due to the low Signal-to-Noise Ratio (SNR). This paper proposes a subject-dependent preprocessing approach that uses Surface Laplacian Filtering and Independent Component Analysis algorithms to remove signal artifacts and improve the classification performance in subjects with poor motor skills.
Article
Engineering, Biomedical
Jingfeng Bi, Ming Chu
Summary: The goal of this study is to design a single-limb, multi-category motor imagery paradigm and achieve cross-subject intention recognition through the transfer data learning network (TDLNet). The network processes cross-subject EEG signals and assigns weights to signal channels using the Residual Attention Mechanism Module (RAMM), resulting in the best classification results.
IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING
(2023)
Article
Engineering, Biomedical
Lie Yang, Yonghao Song, Ke Ma, Enze Su, Longhan Xie
Summary: The proposed motor imagery EEG decoding method based on feature separation effectively improves the decoding accuracy by separating class-related features and class-independent features using the FSNAL network. Experimental results show that the method outperforms other state-of-the-art methods on public EEG datasets, demonstrating its potential for improving the performance of motor imagery BCI systems in the future.
JOURNAL OF NEURAL ENGINEERING
(2021)
Article
Engineering, Biomedical
Do-Yeun Lee, Ji-Hoon Jeong, Byeong-Hoo Lee, Seong-Whan Lee
Summary: The study focused on decoding various forearm movements from EEG signals using a small number of samples. A convolutional neural network based on inter-task transfer learning was proposed, achieving improved classification performance by training the reconstructed ME-EEG signals together with a small amount of MI-EEG signals. The proposed method showed increased performance compared to conventional models, suggesting the feasibility of BCI learning strategies with stable performance using a small calibration dataset and time.
IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING
(2022)
Article
Engineering, Biomedical
Ang Li, Zhenyu Wang, Xi Zhao, Tianheng Xu, Ting Zhou, Honglin Hu
Summary: In this paper, a multi-direction transfer learning strategy is proposed for cross-subject MI EEG-based BCI. This strategy utilizes data from multi-source domains to the target domain as well as from one multi-source domain to another multi-source domain. The strategy is model-independent and can be quickly deployed on existing models, significantly improving model performance and reducing preparation time.
IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING
(2023)
Article
Neurosciences
Kishor Lakshminarayanan, Rakshit Shah, Sohail R. Daulat, Viashen Moodley, Yifei Yao, Deepa Madathil
Summary: This study investigated the effects of combining virtual reality (VR) and action observation on brain activity during motor imagery. The results indicate that combining VR-based action observation enhances brain rhythmic patterns and improves task differentiation compared to motor imagery without action observation.
FRONTIERS IN NEUROSCIENCE
(2023)
Review
Computer Science, Artificial Intelligence
Dongrui Wu, Xue Jiang, Ruimin Peng
Summary: A brain-computer interface (BCI) allows users to communicate with external devices using brain signals, and transfer learning (TL) has been widely used in MI-based BCIs to reduce calibration effort and improve utility.
Article
Engineering, Biomedical
Adel Hameed, Rahma Fourati, Boudour Ammar, Amel Ksibi, Ala Saleh Alluhaidan, Mounir Ben Ayed, Hussain Kareem Khleaf
Summary: This article proposes a transformer-based approach for classification of electroencephalography (EEG) signals in motor imagery (MI), enabling communication in Brain-Computer Interface (BCI) systems. The approach utilizes a self-attention mechanism to extract features in the temporal and spatial domains, improving spatial correlations and classification accuracy. The method has been tested in the BCI Competition IV 2a and 2b benchmarks, outperforming state-of-the-art methods and demonstrating superior stability in subject-dependent and subject-independent strategies.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2024)
Review
Chemistry, Analytical
Arrigo Palumbo, Vera Gramigna, Barbara Calabrese, Nicola Ielpo
Summary: Innovative aids, devices, and assistive technologies are needed to help individuals with severe disabilities live independently and improve overall health. Brain-Computer Interfaces using EEG data show potential for enhancing wheelchair control and movement in people with significant health challenges.
Article
Chemistry, Analytical
Theerat Saichoo, Poonpong Boonbrahm, Yunyong Punsawad
Summary: This study focuses on the use of electroencephalography (EEG) and brain-computer interface (BCI) technology for wheelchair control. By designing tasks and protocols, the study aims to extract efficient control signals from individual users' EEG features. The proposed system achieves high accuracy in command translation and real-time control, providing a potential solution for severe disabilities.
Article
Neurosciences
Gan Huang, Zhiheng Zhao, Shaorong Zhang, Zhenxing Hu, Jiaming Fan, Meisong Fu, Jiale Chen, Yaqiong Xiao, Jun Wang, Guo Dan
Summary: In this study, an online platform for motor-imagery BCI decoding was built to analyze EEG signals from both multi-subject and multi-session experiments. The findings revealed that the time-frequency response of within-subject EEG signals in multi-session experiments was more consistent compared to cross-subject results in multi-subject experiments. Additionally, different strategies for training sample selection should be applied for cross-subject and cross-session tasks. These findings deepen our understanding of inter- and intra-subject variability and can guide the development of new transfer learning methods in EEG-based BCI.
FRONTIERS IN NEUROSCIENCE
(2023)
Article
Computer Science, Artificial Intelligence
D. Collazos-Huertas, D. Cardenas-Pena, G. Castellanos-Dominguez
INTERNATIONAL JOURNAL OF NEURAL SYSTEMS
(2019)
Article
Chemistry, Multidisciplinary
Cristian Torres-Valencia, Alvaro Orozco, David Cardenas-Pena, Andres Alvarez-Meza, Mauricio Alvarez
APPLIED SCIENCES-BASEL
(2020)
Article
Chemistry, Analytical
Andres F. Valencia-Duque, David A. Cardenas-Pena, Andres M. Alvarez-Meza, Alvaro A. Orozco-Gutierrez, Hector F. Quintero-Riaza
Summary: This study introduces a method based on time-delay neural networks to estimate the in-cylinder pressure of a single-cylinder internal combustion engine. Through experiments, it is demonstrated that this method can predict pressure with an accuracy of R2 >0.9 without the need for complicated preprocessing steps.
Article
Chemistry, Multidisciplinary
Ivan De la Pava Panche, Andres Alvarez-Meza, Paula Marcela Herrera Gomez, David Cardenas-Pena, Jorge Ivan Rios Patino, Alvaro Orozco-Gutierrez
Summary: The study introduces a novel method for estimating transfer entropy between neural oscillations by combining a kernel-based estimator with relevance analysis, resulting in an effective connectivity representation that supports classification stages in EEG-based brain-computer interface systems.
APPLIED SCIENCES-BASEL
(2021)
Article
Chemistry, Analytical
Cristian Alfonso Jimenez-Castano, Andres Marino alvarez-Meza, Oscar David Aguirre-Ospina, David Augusto Cardenas-Pena, alvaro Angel Orozco-Gutierrez
Summary: This study introduces a kernel-based deep learning enhancement for nerve structure segmentation, utilizing random Fourier features to improve nerve segmentation. Results show that this method provides better generalization capability for image segmentation of various nerve structures, and GradCam++ is used for data interpretability analysis.
Article
Chemistry, Multidisciplinary
Ivan De La Pava Panche, Viviana Gomez-Orozco, Andres Alvarez-Meza, David Cardenas-Pena, Alvaro Orozco-Gutierrez
Summary: Cross-frequency interactions are important for integrating distributed information in the brain. Using transfer entropy (TE) is a promising approach for estimating these interactions, especially for detecting directed phase-amplitude interactions. By utilizing a kernel-based TE estimator, the proposed method showed improved robustness to noise and the ability to differentiate cognitive load levels in a working memory task based on phase-amplitude interactions.
APPLIED SCIENCES-BASEL
(2021)
Article
Chemistry, Analytical
Yeison Nolberto Cardona-Alvarez, Andres Marino Alvarez-Meza, David Augusto Cardenas-Pena, German Albeiro Castano-Duque, German Castellanos-Dominguez
Summary: This paper presents a flexible and scalable OpenBCI framework for EEG data experiments using the Cyton acquisition board with updated drivers. The framework supports distributed computing tasks, multiple sampling rates, communication protocols, free electrode placement, and single marker synchronization. It provides real-time feedback and controlled execution of EEG-based clinical protocols, along with automatic background configuration and user-friendly widgets for stimuli delivery. The framework offers a promising solution for tailored neurophysiological data processing.
Article
Chemistry, Analytical
Diego Fabian Collazos-Huertas, Andres Marino Alvarez-Meza, David Augusto Cardenas-Pena, German Albeiro Castano-Duque, Cesar German Castellanos-Dominguez
Summary: Motor Imagery (MI) involves imagining motor movements without physical action and has potential applications in rehabilitation and education. The most promising approach for implementing MI is the Brain-Computer Interface (BCI). However, decoding brain responses recorded by scalp electrodes is challenging due to limitations in spatial resolution and non-stationarity. This study proposes a Convolutional Neural Network-based framework to distinguish between MI tasks and identify subjects with poor motor performance at the early stages of BCI training. The proposed method achieves an average accuracy enhancement of 10% compared to the baseline approach, reducing the number of subjects with poor skill from 40% to 20%.
Proceedings Paper
Computer Science, Artificial Intelligence
Julian D. Pastrana-Cortes, Maria Camila Maya-Piedrahita, Paula Marcela Herrera-Gomez, David Cardenas-Pena, Alvaro A. Orozco-Gutierrez
Summary: ADHD, originating in childhood but potentially persisting into adulthood and affecting social skills, presents a diagnostic challenge relying on clinical observation, information from parents and teachers, and clinical expertise. This study proposes a methodology for ADHD diagnosis by extracting features from EEG signals, aiming to address the non-stationarity and non-linearity characteristics hindering the development of diagnostic tools.
PROGRESS IN ARTIFICIAL INTELLIGENCE AND PATTERN RECOGNITION
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Harold Mauricio Diaz-Vargas, Cristian Alfonso Jimenez-Castano, David Augusto Cardenas-Pena, Oscar David Aguirre-Ospina, Alvaro Angel Orozco-Gutierrez
Summary: Peripheral Nerve Blocking (PNB) is a regional anesthesia procedure commonly guided by ultrasound images. An automatic nerve segmentation system can assist specialists in performing successful nerve blocks. The proposed deep neural network, C-UNet, outperforms conventional methods in ultrasound-guided regional anesthesia.
PROGRESS IN ARTIFICIAL INTELLIGENCE AND PATTERN RECOGNITION
(2021)
Proceedings Paper
Acoustics
S. Galindo-Norena, D. Cardenas-Pena, A. A. Orozco-Gutierrez
Summary: ADHD is a childhood-onset neurological disorder that affects attention, memory, and productivity. A feature extraction approach based on EEG signals was proposed to support ADHD diagnosis, showing superior diagnostic capability compared to conventional biomarkers in experimental validation.
28TH EUROPEAN SIGNAL PROCESSING CONFERENCE (EUSIPCO 2020)
(2021)
Proceedings Paper
Acoustics
M. C. Maya-Piedrahita, D. Cardenas-Pena, A. A. Orozco-Gutierrez
Summary: ADHD diagnosis relies on clinical observation and related information, proposing a method to support ADHD diagnosis through EEG characterization. By training HMM and using PPK to measure similarity between patients, support vector machine is used as a diagnostic tool for classification tasks.
28TH EUROPEAN SIGNAL PROCESSING CONFERENCE (EUSIPCO 2020)
(2021)
Article
Mathematics, Interdisciplinary Applications
Luis Ariosto Serna Cardona, Hernan Dario Vargas-Cardona, Piedad Navarro Gonzalez, David Augusto Cardenas Pena, Alvaro Angel Orozco Gutierrez