pith. machine review for the scientific record. sign in

arxiv: 2601.00573 · v2 · submitted 2026-01-02 · 💻 cs.NE · cs.CE

Recognition: no theorem link

Benchmarking ERP Analysis: Manual Features, Deep Learning, and Foundation Models

Yihe Wang , Zhiqiao Kang , Bohan Chen , Yu Zhang , Xiang Zhang
Authors on Pith no claims yet

Pith reviewed 2026-05-16 18:32 UTC · model grok-4.3

classification 💻 cs.NE cs.CE
keywords ERPEEGdeep learningfoundation modelsbenchmarkstimulus classificationdisease detectionTransformer
0
0 comments X

The pith

Benchmark study compares manual features, deep learning, and foundation models for ERP analysis.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper systematically benchmarks traditional manual feature extraction with linear classifiers against deep learning models and pre-trained EEG foundation models for analyzing event-related potentials in EEG signals. It applies a unified preprocessing and training pipeline to evaluate performance on stimulus classification and brain disease detection tasks using 12 publicly available datasets. The work also explores various token-embedding strategies in Transformer-based architectures to find designs suited to ERP characteristics. This comparison aims to establish a practical framework that helps researchers select appropriate methods and design better models for ERP analysis.

Core claim

Through a standardized evaluation of manual, deep learning, and foundation model approaches on representative ERP tasks across multiple datasets, the study provides a landmark framework to guide method selection and tailored model design for future ERP analysis.

What carries the argument

A unified data preprocessing and training pipeline that standardizes comparisons, along with investigations into token-embedding strategies within Transformer architectures for ERP data.

If this is right

  • Researchers gain concrete guidance on when to use manual features versus deep learning or foundation models for ERP tasks.
  • Optimized token embeddings can enhance the performance of Transformer models on ERP signals.
  • The framework enables consistent evaluation and reduces ad-hoc method choices in ERP studies.
  • Findings support the development of more effective models for cognitive analysis and neurological disease detection using ERP.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Applying this benchmarking approach to additional EEG signal types could reveal similar patterns in method effectiveness.
  • Foundation models might particularly benefit ERP analysis in scenarios with limited labeled data.
  • Future work could test the framework's recommendations on real-time or clinical ERP applications.
  • The emphasis on unified pipelines suggests broader standardization efforts in EEG research could improve reproducibility.

Load-bearing premise

The 12 public datasets and two tasks chosen are representative enough of the wider variety of ERP paradigms and clinical applications.

What would settle it

If evaluations on additional ERP datasets or tasks produce substantially different performance rankings or method preferences than those reported in the benchmark.

read the original abstract

Event-related potential (ERP), a specialized paradigm of electroencephalographic (EEG), reflects neurological responses to external stimuli or events, generally associated with the brain's processing of specific cognitive tasks. ERP plays a critical role in cognitive analysis, the detection of neurological diseases, and the assessment of psychological states. Recent years have seen substantial advances in deep learning-based methods for spontaneous EEG and other non-time-locked task-related EEG signals. However, their effectiveness on ERP data remains underexplored, and many existing ERP studies still rely heavily on manually extracted features. In this paper, we conduct a comprehensive benchmark study that systematically compares traditional manual features (followed by a linear classifier), deep learning models, and pre-trained EEG foundation models for ERP analysis. We establish a unified data preprocessing and training pipeline and evaluate these approaches on two representative tasks, ERP stimulus classification and ERP-based brain disease detection, across 12 publicly available datasets. Furthermore, we investigate various token-embedding strategies within advanced Transformer architectures to identify embedding designs that better suit ERP data. Our study provides a landmark framework to guide method selection and tailored model design for future ERP analysis. The code is available at https://github.com/DL4mHealth/ERP-Benchmark

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper benchmarks manual feature extraction followed by linear classifiers against deep learning models and pre-trained EEG foundation models for ERP analysis. It applies a unified preprocessing and training pipeline to two tasks—stimulus classification and disease detection—across 12 public datasets, examines token-embedding strategies inside Transformer architectures, and concludes that the study supplies a landmark framework to guide method selection and tailored model design, with code released at the provided GitHub link.

Significance. If the performance orderings hold under the unified pipeline, the work supplies a reproducible empirical reference that can inform whether manual features, DL, or foundation models are preferable for specific ERP use cases, while the code release and fixed preprocessing strengthen the ability of others to replicate or extend the comparisons.

major comments (2)
  1. [Abstract] Abstract: the claim that the study 'provides a landmark framework to guide method selection' is load-bearing for the paper's contribution, yet the abstract supplies no information on hyperparameter search ranges or the statistical tests used to establish that reported differences are significant; without these details the comparative claims cannot be fully evaluated.
  2. [Abstract] Abstract and dataset-selection section: the two tasks and 12 datasets are asserted to be representative, but the manuscript does not demonstrate that they capture the range of ERP paradigms (e.g., auditory oddball, N400, motor-related potentials), trial counts, channel montages, or artifact profiles typical in the broader literature; this directly affects whether the observed ordering of manual vs. DL vs. foundation-model performance can support general design recommendations.
minor comments (1)
  1. [Abstract] The abstract could state the exact number of datasets and tasks in the opening sentence for immediate clarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on our manuscript. We address each major comment point by point below and have revised the manuscript to incorporate clarifications where appropriate.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that the study 'provides a landmark framework to guide method selection' is load-bearing for the paper's contribution, yet the abstract supplies no information on hyperparameter search ranges or the statistical tests used to establish that reported differences are significant; without these details the comparative claims cannot be fully evaluated.

    Authors: We agree that the abstract would benefit from brief methodological context to support the contribution claim. In the revised manuscript, we will update the abstract to note that hyperparameter optimization was performed via grid search over model-specific ranges (detailed in Section 3.3) and that performance differences were evaluated for statistical significance using paired t-tests with multiple-comparison correction. These procedures are already described in the Methods but their mention in the abstract will make the comparative claims more self-contained. revision: yes

  2. Referee: [Abstract] Abstract and dataset-selection section: the two tasks and 12 datasets are asserted to be representative, but the manuscript does not demonstrate that they capture the range of ERP paradigms (e.g., auditory oddball, N400, motor-related potentials), trial counts, channel montages, or artifact profiles typical in the broader literature; this directly affects whether the observed ordering of manual vs. DL vs. foundation-model performance can support general design recommendations.

    Authors: We concur that explicit demonstration of dataset diversity strengthens the generalizability of the results. In the revised version, we will add a dedicated table and short discussion in the dataset-selection section that enumerates each of the 12 datasets by ERP paradigm (covering auditory oddball, N400, motor-related potentials, and others), trial counts, channel montages, and typical artifact profiles. This addition will directly illustrate the range captured and support the applicability of our method-selection recommendations. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical benchmarks on held-out public data

full rationale

The paper conducts a systematic empirical comparison of manual features, deep learning models, and pre-trained foundation models on 12 public ERP datasets for two tasks (stimulus classification and disease detection). All reported results are performance metrics evaluated on held-out test sets under a unified preprocessing and training pipeline. No equations, derivations, or predictions are present that reduce to fitted parameters or self-referential definitions inside the paper. The landmark-framework claim is an interpretive summary of observed orderings rather than a mathematical result forced by construction. Self-citations, if any, are not load-bearing for the central empirical findings.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The benchmark rests on standard supervised-learning assumptions (i.i.d. train/test splits, cross-entropy loss, standard EEG preprocessing) plus the implicit claim that the chosen 12 datasets adequately sample the ERP domain. No new entities are postulated and no parameters are fitted to produce the headline comparison.

axioms (2)
  • domain assumption Standard EEG preprocessing steps (filtering, epoching, artifact rejection) do not materially alter relative model rankings.
    Invoked when establishing the unified pipeline.
  • domain assumption The 12 public datasets are representative of typical ERP stimulus and clinical tasks.
    Required for the claim that the benchmark guides future method selection.

pith-pipeline@v0.9.0 · 5524 in / 1387 out tokens · 35355 ms · 2026-05-16T18:32:35.853943+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. SCOPE: Structured Prototype-Guided Adaptation for EEG Foundation Models with Limited Labels

    cs.LG 2026-02 unverdicted novelty 6.0

    SCOPE uses cohort-level external supervision, confidence-aware pseudo-labels, and a lightweight prototype-conditioned adapter (ProAdapter) to adapt frozen EEG foundation models in label-limited settings, reporting con...

Reference graph

Works this paper leans on

80 extracted references · 80 canonical work pages · cited by 1 Pith paper · 4 internal anchors

  1. [1]

    Event- related potential studies of emotion regulation: A review of recent progress and future directions

    Annmarie MacNamara, Keanan Joyner, and Julia Klawohn. Event- related potential studies of emotion regulation: A review of recent progress and future directions. International Journal of Psychophysi- ology, 176:73–88, 2022

    work page 2022

  2. [2]

    Evoked and event-related potentials as biomarkers of consciousness state and recovery

    Estelle Pruvost-Robieux, Angela Marchi, Ilaria Martinelli, Eléonore Bouchereau, and Martine Gavaret. Evoked and event-related potentials as biomarkers of consciousness state and recovery. Journal of Clinical Neurophysiology, 39(1):22–31, 2022

    work page 2022

  3. [3]

    Recording and interpreting event-related potentials

    Truett Allison. Recording and interpreting event-related potentials. In Cognitive psychophysiology: Event-related potentials and the study of cognition, pages 1–36. Routledge, 2022

    work page 2022

  4. [4]

    Elec- troencephalography (eeg) and event-related potentials (erps) with human participants

    Gregory A Light, Lisa E Williams, Falk Minow, Joyce Sprock, Anthony Rissling, Richard Sharp, Neal R Swerdlow, and David L Braff. Elec- troencephalography (eeg) and event-related potentials (erps) with human participants. Current protocols in neuroscience , 52(1):6–25, 2010

    work page 2010

  5. [5]

    Event-related potentials

    Steven J Luck. Event-related potentials. 2012

    work page 2012

  6. [6]

    Cognitive neurophysiology: Event-related potentials

    Randolph F Helfrich and Robert T Knight. Cognitive neurophysiology: Event-related potentials. Handbook of clinical neurology , 160:543–558, 2019

    work page 2019

  7. [7]

    Methods for acquiring and analyzing infant event-related potentials

    Tracy DeBoer, Lisa S Scott, and Charles A Nelson. Methods for acquiring and analyzing infant event-related potentials. In Infant EEG and event-related potentials , pages 5–38. Psychology Press, 2013

    work page 2013

  8. [8]

    A survey on deep learning-based non-invasive brain signals: recent advances and new frontiers

    Xiang Zhang, Lina Y ao, Xianzhi Wang, Jessica Monaghan, David Mcalpine, and Y u Zhang. A survey on deep learning-based non-invasive brain signals: recent advances and new frontiers. Journal of neural engineering, 18(3):031002, 2021

    work page 2021

  9. [9]

    Event-related potential: An overview

    Shravani Sur and Vinod Kumar Sinha. Event-related potential: An overview. Industrial psychiatry journal , 18(1):70–73, 2009

    work page 2009

  10. [10]

    Motor imagery eeg signal classification using novel deep learning algorithm

    Sathish Mathiyazhagan and MS Geetha Devasena. Motor imagery eeg signal classification using novel deep learning algorithm. Scientific Reports, 15(1):24539, 2025

    work page 2025

  11. [11]

    Neuript: Foundation model for neural interfaces

    Zitao Fang, Chenxuan Li, Hongting Zhou, Shuyang Y u, Guodong Du, Ashwaq Qasem, Y ang Lu, Jing Li, Junsong Zhang, and Sim Kuan Goh. Neuript: Foundation model for neural interfaces. 39th International Conference on Neural Information Processing Systems , 2025. YIHE et al.: ERP BENCHMARK STUDY 11

    work page 2025

  12. [12]

    Fapex: Fractional amplitude-phase expressor for robust cross-subject seizure prediction

    Ruizhe Zheng, Lingyan Mao, Dingding Han, Tian Luo, Yi Wang, Jing Ding, and Y uguo Y u. Fapex: Fractional amplitude-phase expressor for robust cross-subject seizure prediction. 39th International Conference on Neural Information Processing Systems , 2025

    work page 2025

  13. [13]

    Adformer: A multi-granularity spatial-temporal transformer for eeg-based alzheimer detection

    Yihe Wang, Nadia Mammone, Darina Petrovsky, Alexandros T Tzallas, Francesco C Morabito, and Xiang Zhang. Adformer: A multi-granularity spatial-temporal transformer for eeg-based alzheimer detection. arXiv preprint arXiv:2409.00032 , 2024

    work page arXiv 2024

  14. [14]

    GPT-4 Technical Report

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. Gpt-4 technical report. arXiv preprint arXiv:2303.08774 , 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  15. [15]

    Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context

    Gemini Team, Petko Georgiev, Ving Ian Lei, Ryan Burnell, Libin Bai, Anmol Gulati, Garrett Tanzer, Damien Vincent, Zhufeng Pan, Shibo Wang, et al. Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context. arXiv preprint arXiv:2403.05530 , 2024

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  16. [16]

    DINOv2: Learning Robust Visual Features without Supervision

    Maxime Oquab, Timothée Darcet, Théo Moutakanni, Huy V o, Marc Szafraniec, V asil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, et al. Dinov2: Learning robust visual features without supervision. arXiv preprint arXiv:2304.07193 , 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  17. [17]

    Large brain model for learning generic representations with tremendous eeg data in bci.arXiv preprint arXiv:2405.18765, 2024

    Wei-Bang Jiang, Li-Ming Zhao, and Bao-Liang Lu. Large brain model for learning generic representations with tremendous eeg data in bci. arXiv preprint arXiv:2405.18765 , 2024

    work page arXiv 2024

  18. [18]

    Eegpt: Pretrained transformer for universal and reliable rep- resentation of eeg signals

    Guangyu Wang, Wenchao Liu, Y uhong He, Cong Xu, Lin Ma, and Haifeng Li. Eegpt: Pretrained transformer for universal and reliable rep- resentation of eeg signals. Advances in Neural Information Processing Systems, 37:39249–39280, 2024

    work page 2024

  19. [19]

    Cbramod: A criss-cross brain foundation model for eeg decoding

    Jiquan Wang, Sha Zhao, Zhiling Luo, Y angxuan Zhou, Haiteng Jiang, Shijian Li, Tao Li, and Gang Pan. Cbramod: A criss-cross brain foundation model for eeg decoding. arXiv preprint arXiv:2412.07236 , 2024

    work page arXiv 2024

  20. [20]

    Enhancing motor imagery eeg signal decoding through machine learning: A systematic review of recent progress

    Ibtehaaj Hameed, Danish M Khan, Syed Muneeb Ahmed, Syed Sabeeh Aftab, and Hammad Fazal. Enhancing motor imagery eeg signal decoding through machine learning: A systematic review of recent progress. Computers in Biology and Medicine , 185:109534, 2025

    work page 2025

  21. [21]

    Eeg window length evaluation for the detection of alzheimers disease over different brain regions

    Katerina D Tzimourta, Nikolaos Giannakeas, Alexandros T Tzallas, Loukas G Astrakas, Theodora Afrantou, Panagiotis Ioannidis, Nikolaos Grigoriadis, Pantelis Angelidis, Dimitrios G Tsalikakis, and Markos G Tsipouras. Eeg window length evaluation for the detection of alzheimers disease over different brain regions. Brain sciences , 9(4):81, 2019

    work page 2019

  22. [22]

    Analysis of electroencephalographic signals complexity regarding alzheimer’s disease

    Katerina D Tzimourta, Theodora Afrantou, Panagiotis Ioannidis, Maria Karatzikou, Alexandros T Tzallas, Nikolaos Giannakeas, Loukas G Astrakas, Pantelis Angelidis, Evripidis Glavas, Nikolaos Grigoriadis, et al. Analysis of electroencephalographic signals complexity regarding alzheimer’s disease. Computers & Electrical Engineering , 76:198–212, 2019

    work page 2019

  23. [23]

    Extracting salient features for eeg-based diagnosis of alzheimer’s disease using support vector machine classifier

    NN Kulkarni and VK Bairagi. Extracting salient features for eeg-based diagnosis of alzheimer’s disease using support vector machine classifier. IETE Journal of Research , 63(1):11–22, 2017

    work page 2017

  24. [24]

    Eeg correlates of p300-based brain–computer interface (bci) performance in people with amyotrophic lateral sclerosis

    Joseph N Mak, Dennis J McFarland, Theresa M V aughan, Lynn M Mc- Cane, Phillippa Z Tsui, Debra J Zeitlin, Eric W Sellers, and Jonathan R Wolpaw. Eeg correlates of p300-based brain–computer interface (bci) performance in people with amyotrophic lateral sclerosis. Journal of neural engineering , 9(2):026014, 2012

    work page 2012

  25. [25]

    Identifying patients with poststroke mild cognitive impairment by pattern recognition of working memory load-related erp

    Xiaoou Li, Y uning Y an, and Wenshi Wei. Identifying patients with poststroke mild cognitive impairment by pattern recognition of working memory load-related erp. Computational and Mathematical Methods in Medicine, 2013(1):658501, 2013

    work page 2013

  26. [26]

    Eeg/erp: Within episodic assess- ment framework for cognition

    Bruce Wallace, Frank Knoefel, Rafik Goubran, Rocío A López Zunini, Zhaofen Ren, and Aaron Maccosham. Eeg/erp: Within episodic assess- ment framework for cognition. IEEE Transactions on Instrumentation and Measurement , 66(10):2525–2534, 2017

    work page 2017

  27. [27]

    Characterizing alzheimers disease severity via resting-awake eeg amplitude modulation analysis

    Francisco J Fraga, Tiago H Falk, Paulo AM Kanda, and Renato Anghinah. Characterizing alzheimers disease severity via resting-awake eeg amplitude modulation analysis. PloS one , 8(8):e72240, 2013

    work page 2013

  28. [28]

    Network substrates of cognitive im- pairment in alzheimers disease

    Luke Tait, George Stothart, Elizabeth Coulthard, Jon T Brown, Nina Kazanina, and Marc Goodfellow. Network substrates of cognitive im- pairment in alzheimers disease. Clinical Neurophysiology, 130(9):1581– 1595, 2019

    work page 2019

  29. [29]

    Quantifying synchrony patterns in the eeg of alzheimers patients with linear and non-linear connectivity markers

    Markus Waser, Heinrich Garn, Reinhold Schmidt, Thomas Benke, Peter Dal-Bianco, Gerhard Ransmayr, Helena Schmidt, Stephan Seiler, Günter Sanin, Florian Mayer, et al. Quantifying synchrony patterns in the eeg of alzheimers patients with linear and non-linear connectivity markers. Journal of Neural Transmission , 123:297–316, 2016

    work page 2016

  30. [30]

    Improving alzheimer’s disease diagnosis with machine learning techniques

    Lucas R Trambaiolli, Ana C Lorena, Francisco J Fraga, Paulo AM Kanda, Renato Anghinah, and Ricardo Nitrini. Improving alzheimer’s disease diagnosis with machine learning techniques. Clinical EEG and neuroscience, 42(3):160–165, 2011

    work page 2011

  31. [31]

    Index of alpha/theta ratio of the electroencephalogram: a new marker for alzheimers disease

    Magali T Schmidt, Paulo AM Kanda, Luis FH Basile, Helder Frederico da Silva Lopes, Regina Baratho, Jose LC Demario, Mario S Jorge, Antonio E Nardi, Sergio Machado, Jéssica N Ianof, et al. Index of alpha/theta ratio of the electroencephalogram: a new marker for alzheimers disease. Frontiers in aging neuroscience , 5:60, 2013

    work page 2013

  32. [32]

    Multiple characteristics analysis of alzheimers electroencephalogram by power spectral density and lempel–ziv complexity

    Xiaokun Liu, Chunlai Zhang, Zheng Ji, Yi Ma, Xiaoming Shang, Qi Zhang, Wencheng Zheng, Xia Li, Jun Gao, Ruofan Wang, et al. Multiple characteristics analysis of alzheimers electroencephalogram by power spectral density and lempel–ziv complexity. Cognitive neurody- namics, 10:121–133, 2016

    work page 2016

  33. [33]

    Clinicians road map to wavelet eeg as an alzheimers disease biomarker

    Paulo Afonso Medeiros Kanda, Lucas R Trambaiolli, Ana C Lorena, Francisco J Fraga, Luis Fernando I Basile, Ricardo Nitrini, and Renato Anghinah. Clinicians road map to wavelet eeg as an alzheimers disease biomarker. Clinical EEG and neuroscience , 45(2):104–112, 2014

    work page 2014

  34. [34]

    Quantitative eeg markers relate to alzheimers disease severity in the prospective dementia registry austria (prodem)

    Heinrich Garn, Markus Waser, Manfred Deistler, Thomas Benke, Peter Dal-Bianco, Gerhard Ransmayr, Helena Schmidt, Guenter Sanin, Peter Santer, Georg Caravias, et al. Quantitative eeg markers relate to alzheimers disease severity in the prospective dementia registry austria (prodem). Clinical Neurophysiology, 126(3):505–513, 2015

    work page 2015

  35. [35]

    Multiscale fluctuation-based dispersion entropy and its applications to neurological diseases

    Hamed Azami, Steven E Arnold, Saeid Sanei, Zhuoqing Chang, Guillermo Sapiro, Javier Escudero, and Anoopum S Gupta. Multiscale fluctuation-based dispersion entropy and its applications to neurological diseases. IEEE Access , 7:68718–68733, 2019

    work page 2019

  36. [36]

    Unbiased estimation of permutation entropy in eeg analysis for alzheimer’s disease classification

    Lucie Tylová, Jaromír Kukal, Václav Hubata-V acek, and Old ˇrich Vyšata. Unbiased estimation of permutation entropy in eeg analysis for alzheimer’s disease classification. Biomedical Signal Processing and Control, 39:424–430, 2018

    work page 2018

  37. [37]

    Early screening of children with autism spectrum disorder based on elec- troencephalogram signal feature selection with l1-norm regularization

    Shixin Peng, Ruyi Xu, Xin Yi, Xin Hu, Lili Liu, and Leyuan Liu. Early screening of children with autism spectrum disorder based on elec- troencephalogram signal feature selection with l1-norm regularization. Frontiers in Human Neuroscience , 15:656578, 2021

    work page 2021

  38. [38]

    Eeg-based affective state recognition from human brain signals by using hjorth-activity

    Raja Majid Mehmood, Muhammad Bilal, S Vimal, and Seong-Whan Lee. Eeg-based affective state recognition from human brain signals by using hjorth-activity. Measurement, 202:111738, 2022

    work page 2022

  39. [39]

    Diagnosis of alzheimers disease via machine learning approaches with integrated resting-state eeg and erp characteristics

    Zhangjing Deng, Haoying Bai, Shijing Wu, Jiani Wu, Boyuan Xia, Yingxi Chen, Y urou He, Shuyu Li, and Y ang Lü. Diagnosis of alzheimers disease via machine learning approaches with integrated resting-state eeg and erp characteristics. Cognitive Computation , 17(6):166, 2025

    work page 2025

  40. [40]

    Eegnet: a compact convolutional neural network for eeg-based brain–computer interfaces

    V ernon J Lawhern, Amelia J Solon, Nicholas R Waytowich, Stephen M Gordon, Chou P Hung, and Brent J Lance. Eegnet: a compact convolutional neural network for eeg-based brain–computer interfaces. Journal of neural engineering , 15(5):056013, 2018

    work page 2018

  41. [41]

    Eeg-inception: a novel deep convolutional neural network for assistive erp-based brain-computer interfaces

    Eduardo Santamaria-V azquez, Victor Martinez-Cagigal, Fernando V aquerizo-Villar, and Roberto Hornero. Eeg-inception: a novel deep convolutional neural network for assistive erp-based brain-computer interfaces. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 28(12):2773–2782, 2020

    work page 2020

  42. [42]

    An attention-based wavelet convolution neural network for epilepsy eeg classification

    Qi Xin, Shaohai Hu, Shuaiqi Liu, Ling Zhao, and Y u-Dong Zhang. An attention-based wavelet convolution neural network for epilepsy eeg classification. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 30:957–966, 2022

    work page 2022

  43. [43]

    A transformer-based approach combining deep learning network and spatial-temporal information for raw eeg classification

    Jin Xie, Jie Zhang, Jiayao Sun, Zheng Ma, Liuni Qin, Guanglin Li, Huihui Zhou, and Y ang Zhan. A transformer-based approach combining deep learning network and spatial-temporal information for raw eeg classification. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 30:2126–2136, 2022

    work page 2022

  44. [44]

    Eeg conformer model based epileptic seizure prediction using deep learning

    N Kasthuri, R Ramyea, VS Arunprasshath, S Abhineeth, and S Bharathraj. Eeg conformer model based epileptic seizure prediction using deep learning. In 2024 15th International Conference on Com- puting Communication and Networking Technologies (ICCCNT) , pages 1–7. IEEE, 2024

    work page 2024

  45. [45]

    Neuro-bert: Rethink- ing masked autoencoding for self-supervised neurological pretraining

    Di Wu, Siyuan Li, Jie Y ang, and Mohamad Sawan. Neuro-bert: Rethink- ing masked autoencoding for self-supervised neurological pretraining. arXiv preprint arXiv:2204.12440 , 2022

    work page arXiv 2022

  46. [46]

    Lggnet: Learning from local-global-graph representations for brain–computer interface

    Yi Ding, Neethu Robinson, Chengxuan Tong, Qiuhao Zeng, and Cuntai Guan. Lggnet: Learning from local-global-graph representations for brain–computer interface. IEEE Transactions on Neural Networks and Learning Systems , 35(7):9773–9786, 2023

    work page 2023

  47. [47]

    Mocnn: A multiscale deep convolutional neural network for erp-based brain-computer interfaces

    Jing Jin, Ruitian Xu, Ian Daly, Xueqing Zhao, Xingyu Wang, and Andrzej Cichocki. Mocnn: A multiscale deep convolutional neural network for erp-based brain-computer interfaces. IEEE Transactions on Cybernetics , 54(9):5565–5576, 2024

    work page 2024

  48. [48]

    Eegmamba: Bidirectional state space model with mixture of experts for eeg multi-task classification

    Yiyu Gui, MingZhi Chen, Y uqi Su, Guibo Luo, and Y uchao Y ang. Eegmamba: Bidirectional state space model with mixture of experts for eeg multi-task classification. arXiv preprint arXiv:2407.20254 , 2024. 12

    work page arXiv 2024

  49. [49]

    Repurposing foundation model for generalizable medical time series classification

    Nan Huang, Haishuai Wang, Zihuai He, Marinka Zitnik, and Xiang Zhang. Repurposing foundation model for generalizable medical time series classification. arXiv preprint arXiv:2410.03794 , 2024

    work page arXiv 2024

  50. [50]

    Neurolm: A universal multi-task foundation model for bridging the gap between language and eeg signals.arXiv preprint arXiv:2409.00101, 2024

    Wei-Bang Jiang, Y ansen Wang, Bao-Liang Lu, and Dongsheng Li. Neurolm: A universal multi-task foundation model for bridging the gap between language and eeg signals. arXiv preprint arXiv:2409.00101 , 2024

    work page arXiv 2024

  51. [51]

    Luna: Efficient and topology-agnostic foundation model for eeg signal analy- sis

    Berkay Döner, Thorir Mar Ingolfsson, Luca Benini, and Y awei Li. Luna: Efficient and topology-agnostic foundation model for eeg signal analy- sis. 39th International Conference on Neural Information Processing Systems, 2025

    work page 2025

  52. [52]

    Csbrain: A cross-scale spatiotemporal brain foundation model for eeg decoding

    Y uchen Zhou, Jiamin Wu, Zichen Ren, Zhouheng Y ao, Weiheng Lu, Kunyu Peng, Qihao Zheng, Chunfeng Song, Wanli Ouyang, and Chao Gou. Csbrain: A cross-scale spatiotemporal brain foundation model for eeg decoding. 39th International Conference on Neural Information Processing Systems , 2025

    work page 2025

  53. [53]

    Economical assessment of working memory and response inhibition in adhd using a combined n-back/nogo paradigm: An erp study

    Carolin Breitling-Ziegler, Jana Tegelbeckers, Hans-Henning Flechtner, and Kerstin Krauel. Economical assessment of working memory and response inhibition in adhd using a combined n-back/nogo paradigm: An erp study. Frontiers in human neuroscience , 14:322, 2020

    work page 2020

  54. [54]

    Eeg: 3-stim auditory oddball and rest in parkinsons

    JF Cavanagh. Eeg: 3-stim auditory oddball and rest in parkinsons. OpenNeuro, OpenNeuro, 2021

    work page 2021

  55. [55]

    Cognitive electrophysiology in socioeconomic context in adulthood

    Elif Isbell, Amanda N Peters, Dylan M Richardson, and Nancy E Rodas De León. Cognitive electrophysiology in socioeconomic context in adulthood. Scientific Data , 12(1):841, 2025

    work page 2025

  56. [56]

    Erps predict symptomatic distress and recovery in sub-acute mild traumatic brain injury

    James F Cavanagh, J Kevin Wilson, Rebecca E Rieger, Darbi Gill, James M Broadway, Jacqueline Hope Story Remer, Violet Fratzke, Andrew R Mayer, and Davin K Quinn. Erps predict symptomatic distress and recovery in sub-acute mild traumatic brain injury. Neuropsychologia, 132:107125, 2019

    work page 2019

  57. [57]

    The nencki-symfonia electroencephalography/event-related potential dataset: Multiple cognitive tasks and resting-state data collected in a sample of healthy adults

    Patrycja Dzianok, Ingrida Antonova, Jakub Wojciechowski, Joanna Dreszer, and Ewa Kublik. The nencki-symfonia electroencephalography/event-related potential dataset: Multiple cognitive tasks and resting-state data collected in a sample of healthy adults. GigaScience, 11:giac015, 2022

    work page 2022

  58. [58]

    Evoked mid-frontal activity predicts cognitive dysfunction in parkinsons disease

    Arun Singh, Rachel C Cole, Arturo I Espinoza, Jan R Wessel, James F Cavanagh, and Nandakumar S Narayanan. Evoked mid-frontal activity predicts cognitive dysfunction in parkinsons disease. Journal of Neurol- ogy, Neurosurgery & Psychiatry , 94(11):945–953, 2023

    work page 2023

  59. [59]

    An eeg marker of reward processing is diminished in parkinsons disease

    Darin R Brown, Sarah Pirio Richardson, and James F Cavanagh. An eeg marker of reward processing is diminished in parkinsons disease. Brain research, 1727:146541, 2020

    work page 2020

  60. [60]

    Mid-frontal theta activity is diminished during cognitive control in parkinson’s disease

    Arun Singh, Sarah Pirio Richardson, Nandakumar Narayanan, and James F Cavanagh. Mid-frontal theta activity is diminished during cognitive control in parkinson’s disease. Neuropsychologia, 117:113– 122, 2018

    work page 2018

  61. [61]

    Iclabel: An automated electroencephalographic independent component classi- fier, dataset, and website

    Luca Pion-Tonachini, Ken Kreutz-Delgado, and Scott Makeig. Iclabel: An automated electroencephalographic independent component classi- fier, dataset, and website. NeuroImage, 198:181–197, 2019

    work page 2019

  62. [62]

    Enhanced gamma activity and cross-frequency interaction of resting-state electroencephalographic oscillations in patients with alzheimers disease

    Jing Wang, Y uxing Fang, Xiao Wang, Huichao Y ang, Xin Y u, and Huali Wang. Enhanced gamma activity and cross-frequency interaction of resting-state electroencephalographic oscillations in patients with alzheimers disease. Frontiers in aging neuroscience , 9:243, 2017

    work page 2017

  63. [63]

    The effects of automated artifact removal algo- rithms on electroencephalography-based alzheimer’s disease diagnosis

    Raymundo Cassani, Tiago H Falk, Francisco J Fraga, Paulo AM Kanda, and Renato Anghinah. The effects of automated artifact removal algo- rithms on electroencephalography-based alzheimer’s disease diagnosis. Frontiers in aging neuroscience , 6:55, 2014

    work page 2014

  64. [64]

    Multiple feature extraction and classification of electroencephalo- graph signal for alzheimers’ with spectrum and bispectrum

    Ruofan Wang, Jiang Wang, Shunan Li, Haitao Y u, Bin Deng, and Xile Wei. Multiple feature extraction and classification of electroencephalo- graph signal for alzheimers’ with spectrum and bispectrum. Chaos: An Interdisciplinary Journal of Nonlinear Science , 25(1), 2015

    work page 2015

  65. [65]

    Eeg in the diagnostics of alzheimers disease

    Markus Waser, Manfred Deistler, Heinrich Garn, Thomas Benke, Pe- ter Dal-Bianco, Gerhard Ransmayr, Dieter Grossegger, and Reinhold Schmidt. Eeg in the diagnostics of alzheimers disease. Statistical Papers, 54:1095–1107, 2013

    work page 2013

  66. [66]

    Predictive models in diagnosis of alzheimers disease from eeg

    Lucie Tylova, Jaromir Kukal, and Oldrich Vysata. Predictive models in diagnosis of alzheimers disease from eeg. Acta Polytechnica, 53(2), 2013

    work page 2013

  67. [67]

    Scale-free behaviour and metastable brain-state switching driven by human cognition, an empirical approach

    Aldo Mora-Sánchez, Gérard Dreyfus, and François-Benoît Vialatte. Scale-free behaviour and metastable brain-state switching driven by human cognition, an empirical approach. Cognitive neurodynamics , 13:437–452, 2019

    work page 2019

  68. [68]

    Eeg alpha activity and the erp to target stimuli in an auditory oddball paradigm

    Robert J Barry, Sopa Kirkaikul, and Darren Hodder. Eeg alpha activity and the erp to target stimuli in an auditory oddball paradigm. International journal of psychophysiology , 39(1):39–50, 2000

    work page 2000

  69. [69]

    Temporal convolutional networks for action segmentation and detection

    Colin Lea, Michael D Flynn, Rene Vidal, Austin Reiter, and Gregory D Hager. Temporal convolutional networks for action segmentation and detection. In proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 156–165, 2017

    work page 2017

  70. [70]

    Moderntcn: A modern pure convolution structure for general time series analysis

    Donghao Luo and Xue Wang. Moderntcn: A modern pure convolution structure for general time series analysis. In The twelfth international conference on learning representations , pages 1–43, 2024

    work page 2024

  71. [71]

    TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis

    Haixu Wu, Tengge Hu, Y ong Liu, Hang Zhou, Jianmin Wang, and Mingsheng Long. Timesnet: Temporal 2d-variation modeling for general time series analysis. arXiv preprint arXiv:2210.02186 , 2022

    work page internal anchor Pith review arXiv 2022

  72. [72]

    A time series is worth 64words: Long-term forecasting with transformers

    Y Nie. A time series is worth 64words: Long-term forecasting with transformers. International conference on learning representations , 2023

    work page 2023

  73. [73]

    itransformer: Inverted transformers are ef- fective for time series forecasting

    Y ong Liu, Tengge Hu, Haoran Zhang, Haixu Wu, Shiyu Wang, Lintao Ma, and Mingsheng Long. itransformer: Inverted transformers are ef- fective for time series forecasting. International conference on learning representations, 2023

    work page 2023

  74. [74]

    Med- former: A multi-granularity patching transformer for medical time-series classification

    Yihe Wang, Nan Huang, Taida Li, Y ujun Y an, and Xiang Zhang. Med- former: A multi-granularity patching transformer for medical time-series classification. Advances in Neural Information Processing Systems , 37:36314–36341, 2024

    work page 2024

  75. [75]

    Towards multi-resolution spatiotempo- ral graph learning for medical time series classification

    Wei Fan, Jingru Fei, Dingyu Guo, Kun Yi, Xiaozhuang Song, Haolong Xiang, Hangting Y e, and Min Li. Towards multi-resolution spatiotempo- ral graph learning for medical time series classification. In Proceedings of the ACM on Web Conference 2025 , pages 5054–5064, 2025

    work page 2025

  76. [76]

    Xception: Deep learning with depthwise separable convolutions

    François Chollet. Xception: Deep learning with depthwise separable convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 1251–1258, 2017

    work page 2017

  77. [77]

    Eeg conformer: Convolutional transformer for eeg decoding and visu- alization

    Y onghao Song, Qingqing Zheng, Bingchuan Liu, and Xiaorong Gao. Eeg conformer: Convolutional transformer for eeg decoding and visu- alization. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 31:710–719, 2022

    work page 2022

  78. [78]

    Biot: Biosignal transformer for cross-data learning in the wild

    Chaoqi Y ang, M Westover, and Jimeng Sun. Biot: Biosignal transformer for cross-data learning in the wild. Advances in Neural Information Processing Systems , 36:78240–78260, 2023

    work page 2023

  79. [79]

    Lead: Large foundation model for eeg-based alzheimer’s disease detection

    Yihe Wang, Nan Huang, Nadia Mammone, Marco Cecchi, and Xiang Zhang. Lead: Large foundation model for eeg-based alzheimer’s disease detection. arXiv preprint arXiv:2502.01678 , 2025

    work page arXiv 2025

  80. [80]

    The temple university hospital eeg data corpus

    Iyad Obeid and Joseph Picone. The temple university hospital eeg data corpus. Frontiers in neuroscience , 10:196, 2016

    work page 2016