pith. sign in

arxiv: 2606.00106 · v1 · pith:4P3MLPYQnew · submitted 2026-05-26 · 📡 eess.SP · cs.AI· cs.HC· cs.LG

A Methodological Framework for Explicit Control of the Speed-Accuracy Trade-off in Brain-Computer Interfaces

Javier Jim\'enez , Francisco B Rodr\'iguez This is my paper

Pith reviewed 2026-06-29 15:26 UTC · model grok-4.3

classification 📡 eess.SP cs.AIcs.HCcs.LG
keywords brain-computer interfacespeed-accuracy trade-offP300 paradigmGain measureConservation measurealpha parameterevaluation frameworkinformation transfer rate
0
0 comments X

The pith

A tunable alpha parameter in a Gain-Conservation framework gives explicit control over the speed-accuracy trade-off in BCIs without changing the classifier.

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

The paper introduces an evaluation framework that separates speed and accuracy in brain-computer interfaces through two measures: Gain for relative speed improvement and Conservation for relative accuracy preservation. These combine into a Gain-Cons Balance controlled by the parameter alpha, which shifts the operating point toward faster, more accurate, or balanced performance. The framework is shown to work independently of the classifier, paradigm, and early-stopping strategy on P300 data from 63 subjects. Tuning alpha produces distinct behaviors while also exposing biases in the standard Information Transfer Rate metric. The approach supports subject-level predictions and makes BCI behavior more explainable.

Core claim

The central claim is that the Gain-Cons Balance, formed by combining Gain and Conservation and regulated by alpha, functions as a controllable design variable for the speed-accuracy trade-off. This balance adjusts BCI operation to application needs while remaining independent of the underlying classifier, paradigm, and early-stopping strategy, as demonstrated through validation on public P300 recordings with multiple classifiers and strategies.

What carries the argument

The Gain-Cons Balance, a tunable linear combination of relative speed improvement (Gain) and relative accuracy preservation (Conservation) controlled by the scalar parameter alpha.

If this is right

  • Tuning alpha enables application-specific BCI optimization without retraining classifiers.
  • Subject-level performance can be predicted from the framework measurements.
  • The Information Transfer Rate's bias toward speed is explained by the separate Gain and Conservation components.
  • BCI evaluation becomes more transparent across different setups.
  • Distinct operating points in speed, accuracy, and bitrate are achievable by alpha adjustment.

Where Pith is reading between the lines

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

  • The framework could support real-time alpha adaptation during BCI use to match changing task demands.
  • It might extend to other BCI modalities like motor imagery or steady-state visual evoked potentials.
  • Standardized reporting of alpha values could improve comparability across BCI studies.
  • If the measures prove robust, they could replace or augment ITR in BCI benchmarks.

Load-bearing premise

The Gain and Conservation measures, and their alpha-controlled combination, remain independent of the classifier, paradigm, and early-stopping strategy while separating speed from accuracy without new biases.

What would settle it

Re-running the same alpha values on a new classifier or paradigm that produces speed-accuracy points inconsistent with the predicted Gain-Cons curve would falsify the independence claim.

Figures

Figures reproduced from arXiv: 2606.00106 by Francisco B Rodr\'iguez, Javier Jim\'enez.

Figure 1
Figure 1. Figure 1: Closed-loop BCI system composed of a transducer, a control-interface, and a device controller. A grey line highlights BCIs can stop sending commands if not enough evidence is available. Inspired from [4, 20]. This diagram provides a conceptual framework for the separation of signal decoding and decision control. ∗Corresponding author javier.jimenez01@uam.es (J. Jiménez); f.rodriguez@uam.es (F.B. Rodríguez)… view at source ↗
Figure 2
Figure 2. Figure 2: The figure illustrates the combined optimization and validation procedure to adjust both transducer and control￾interface. This process was repeated for every session left out. Results on test were averaged across sessions for every subject. Within our framework, transducers aim to translate brain signals into evidence and are optimized by maximizing the J. Jiménez et al.: Preprint submitted to Elsevier Pa… view at source ↗
Figure 3
Figure 3. Figure 3: presents two-dimensional histograms representing the joint distribution of achieved accuracy and required trials for each optimization policy. Color intensity reflects the empirical likelihood of observing a given speed–accuracy configuration. These distributions are obtained by aggre￾gating all subjects, classifiers, and early-stopping strategies, providing an overall estimate of the system’s behavior. Fo… view at source ↗
Figure 4
Figure 4. Figure 4: (A) Average required trials and obtained accuracies across classifiers, early-stopping strategies, and optimization policies for Hoffmann et al. RSVP dataset [11]; the fitted line illustrates the speed–accuracy trade-off; cross markers indicate the overall mean and standard deviation per optimization policy. (B) Proportion of experiments in which each optimization policy won, tied, or lost against others i… view at source ↗
Figure 5
Figure 5. Figure 5: Quadrants (Q1-Q4) and subquadrants (Q41-Q44) division example for every optimization policy on Hoffmann et al. RSVP dataset [11]. This distinction allows for a structured comparison of speed and accuracy profiles across different optimization strategies. Moreover, the mean and standard deviation of accuracy and required trials were also computed for each classifier, early-stopping strategy, and optimizatio… view at source ↗
Figure 7
Figure 7. Figure 7: A) presents conditional distributions of required trials to achieve at least 80% accuracy for each optimization policy. Crucially, ITR and GCB(𝛼 = 0.75) concentrate most of their probability mass—about 60%—within the first trials, whereas GCB(𝛼 = 0.25) accumulates about 90% of its mass across more trials. GCB(𝛼 = 0.5) shows an intermediate trade-off. Importantly, as just 60% of the ITR and GCB(𝛼 = 0.75) ma… view at source ↗
Figure 8
Figure 8. Figure 8: ITR estimated model and absolute error across different experimental parameters. A) Results for Hoffmann et al. [11] parameters. B) Results for Won et al. [32] parameters. The fitted models illustrate the sensitivity of the ITR to variations in speed and accuracy. The estimated non-linear models explained 96.17% and 97.33% of the ITR variance (𝑅2 ) for Hoffmann et al. [11] J. Jiménez et al.: Preprint submi… view at source ↗
Figure 9
Figure 9. Figure 9: Jensen–Shannon distance between BCIs optimized with the ITR and BCIs adjusted with different GCB’s 𝛼 across RSVP (A), RCP (B), and stimulus-level RCP (C) paradigms. The closer to zero, the more similar and vice versa. This analysis quantifies the empirical similarity between systems optimized using the ITR and those using the GCB at different 𝛼 values, identifying the range of 𝛼 values that best aligns wit… view at source ↗
Figure 10
Figure 10. Figure 10: Reproduced results on the RCP modality from Won et al. dataset [32]. (A) Average required trials and obtained accuracies across classifiers, early-stopping strategies, and optimization policies; the fitted line illustrates the speed–accuracy trade-off; cross markers indicate the overall mean and standard deviation per optimization policy. (B) Proportion of experiments in which each optimization policy won… view at source ↗
Figure 11
Figure 11. Figure 11: Reproduced results on the stimulus-level RCP modality from Won et al. dataset [32]. (A) Average required trials and obtained accuracies across classifiers, early-stopping strategies, and optimization policies; the fitted line illustrates the speed– accuracy trade-off; cross markers indicate the overall mean and standard deviation per optimization policy. (B) Proportion of experiments in which each optimiz… view at source ↗
read the original abstract

Brain-computer interfaces (BCIs) are limited by low signal-to-noise ratio in modalities such as electroencephalography, which requires multiple trials to reliably decode user intentions. This induces a speed-accuracy trade-off, whereby higher accuracy comes at the cost of speed. The speed-accuracy balance is application-dependent, motivating controllable trade-offs. Conventional metrics, such as the Information Transfer Rate, combine speed and accuracy obscuring their dependence and potentially introducing biases. In this study, we propose an evaluation framework independent of classifier, paradigm, and early-stopping strategy that separates speed and accuracy. We employ two measures, Gain (relative speed improvement) and Conservation (relative accuracy preservation), and combine them into a tunable Gain-Cons Balance controlled by {\alpha}, regulating the speed-accuracy trade-off. The parameter adjusts the operating point without modifying the classifier, facilitating deployment across scenarios. The framework was evaluated on P300 event-related potential paradigms using public recordings from 63 subjects as well as multiple classifiers and early-stopping strategies to achieve distinct operating points in speed-accuracy and bitrate. Results show that tuning {\alpha} yields fast, accurate, or balanced BCI behaviours, demonstrating explicit control of the speed-accuracy trade-off. The method supports subject-level performance prediction and improves explainability of BCI behaviour. Further analysis of the Information Transfer Rate reveals a systematic bias toward speed, explained by the proposed framework through the Gain and Conservation measurements. Overall, this work establishes the speed-accuracy trade-off as a controllable design variable validated on public P300-based paradigms, enabling transparent evaluation and application-specific optimization of BCIs.

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

1 major / 0 minor

Summary. The paper proposes a methodological framework for explicit control of the speed-accuracy trade-off in BCIs. It introduces Gain (relative speed improvement) and Conservation (relative accuracy preservation) measures, combined into a tunable Gain-Cons Balance via parameter α. The framework is claimed to be independent of classifier, paradigm, and early-stopping strategy. It is evaluated on P300 ERP data from 63 subjects using multiple classifiers and stopping strategies, showing that α tuning achieves distinct operating points, supports subject-level prediction, and explains ITR biases toward speed.

Significance. If the measures are robustly independent of paradigm and cleanly separate speed from accuracy without new biases, the framework would enable application-specific BCI optimization without classifier retraining and improve explainability over conventional metrics like ITR. The use of public recordings and multiple classifiers strengthens reproducibility.

major comments (1)
  1. [Abstract] Abstract: The manuscript asserts independence from paradigm, yet validation is confined to P300 ERP across all 63 subjects and all reported classifiers/stopping rules. This is load-bearing for the central claim, as Gain (relative speed improvement) and Conservation (relative accuracy preservation) may embed P300-specific trial-structure assumptions that do not generalize to asynchronous paradigms such as motor imagery or SSVEP.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the detailed and constructive review. The major comment raises a valid point about the scope of empirical validation relative to the independence claim. We address it directly below and will revise the manuscript accordingly.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The manuscript asserts independence from paradigm, yet validation is confined to P300 ERP across all 63 subjects and all reported classifiers/stopping rules. This is load-bearing for the central claim, as Gain (relative speed improvement) and Conservation (relative accuracy preservation) may embed P300-specific trial-structure assumptions that do not generalize to asynchronous paradigms such as motor imagery or SSVEP.

    Authors: We agree that the empirical validation is limited to P300 ERP data and that this is a substantive limitation for the paradigm-independence claim. The Gain and Conservation measures are defined generally (as relative speed improvement and accuracy preservation against a no-early-stopping baseline) without explicit dependence on synchronous trial structures; however, we cannot rule out that their numerical behavior may interact with paradigm-specific timing or decision boundaries. To address this, we will revise the manuscript to (1) qualify the independence claim as conceptual and methodological rather than fully empirically demonstrated across paradigms, (2) add a new subsection discussing potential limitations and required adaptations for asynchronous paradigms (motor imagery, SSVEP), and (3) outline a concrete plan for future multi-paradigm validation using public datasets. These changes will be reflected in the abstract, introduction, and discussion. revision: yes

Circularity Check

0 steps flagged

No circularity detected in derivation chain

full rationale

The provided manuscript text defines Gain as relative speed improvement and Conservation as relative accuracy preservation, then combines them via a tunable alpha-weighted balance to control the speed-accuracy trade-off. No equations, self-referential definitions, fitted inputs renamed as predictions, or load-bearing self-citations appear in the abstract or surrounding text that would reduce these measures or the alpha control to the paper's own inputs by construction. The framework is presented as independent of classifier, paradigm, and stopping strategy, with evaluation on public P300 data; the central claims rest on this separation without reducing to tautology or prior self-citation chains. The derivation is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 0 axioms · 0 invented entities

Alpha is introduced as the single tunable control parameter. No other free parameters, axioms, or invented entities are stated in the abstract. The independence from classifier/paradigm is treated as a domain assumption rather than derived.

free parameters (1)
  • alpha
    Tunable parameter that sets the operating point on the speed-accuracy trade-off.

pith-pipeline@v0.9.1-grok · 5835 in / 1125 out tokens · 35648 ms · 2026-06-29T15:26:09.775159+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

35 extracted references · 25 canonical work pages

  1. [1]

    A Bayesian dynamic stop- ping method for evoked response brain-computer interfacing

    Ahmadi, S., Desain, P., Thielen, J., 2024. A Bayesian dynamic stop- ping method for evoked response brain-computer interfacing. Fron- tiers in Human Neuroscience 18. doi:10.3389/fnhum.2024.1437965

    work page doi:10.3389/fnhum.2024.1437965 2024

  2. [2]

    Akiba, T., Sano, S., Yanase, T., Ohta, T., Koyama, M., 2019. Op- tuna: A Next-generation Hyperparameter Optimization Framework, in: Proceedings of the 25th ACM SIGKDD International Conference onKnowledgeDiscovery&DataMining,AssociationforComputing Machinery, New York, NY, USA. pp. 2623–2631. doi:10.1145/ 3292500.3330701

    work page arXiv 2019

  3. [3]

    Information transfer rate in BCIs: Towards tightly integrated symbiosis

    Arslan, S.S., Sinha, P., 2024. Information transfer rate in BCIs: Towards tightly integrated symbiosis. Biomedical Signal Processing and Control 87, 105466. doi:10.1016/j.bspc.2023.105466

    work page doi:10.1016/j.bspc.2023.105466 2024

  4. [4]

    Performances Evaluation and Optimization of Brain ComputerInterfaceSystemsinaCopySpellingTask

    Bianchi,L.,Quitadamo,L.R.,Garreffa,G.,Cardarilli,G.C.,Marciani, M.G., 2007. Performances Evaluation and Optimization of Brain ComputerInterfaceSystemsinaCopySpellingTask. IEEETransac- tionsonNeuralSystemsandRehabilitationEngineering15,207–216. doi:10.1109/TNSRE.2007.897024

    work page doi:10.1109/tnsre.2007.897024 2007

  5. [5]

    Channel capacity in brain–computer interfaces

    Bulhões da Silva Costa, T., Fernanda Suarez Uribe, L., Negreiros de Carvalho, S., Coutinho Soriano, D., Castellano, G., Suyama, R., Attux, R., Panazio, C., 2020. Channel capacity in brain–computer interfaces. Journal of Neural Engineering 17, 016060. doi:10.1088/ 1741-2552/ab6cb7. J. Jiménez et al.:Preprint submitted to ElsevierPage 12 of 13 Framework to ...

    2020

  6. [6]

    A Coefficient of Agreement for Nominal Scales

    Cohen, J., 1960. A Coefficient of Agreement for Nominal Scales. EducationalandPsychologicalMeasurement20,37–46. doi:10.1177/ 001316446002000104

    1960

  7. [7]

    IEEE Transactions on Neural Systems and Rehabilitation Engineering 18, 20–28

    DalSeno,B.,Matteucci,M.,Mainardi,L.T.,2010.TheUtilityMetric: A Novel Method to Assess the Overall Performance of Discrete Brain–Computer Interfaces. IEEE Transactions on Neural Systems and Rehabilitation Engineering 18, 20–28. doi:10.1109/TNSRE.2009. 2032642

    work page doi:10.1109/tnsre.2009 2010

  8. [8]

    Non-Invasive Brain-Computer Interfaces: StateoftheArtandTrends.IEEEReviewsinBiomedicalEngineering 18, 26–49

    Edelman, B.J., Zhang, S., Schalk, G., Brunner, P., Müller-Putz, G., Guan, C., He, B., 2025. Non-Invasive Brain-Computer Interfaces: StateoftheArtandTrends.IEEEReviewsinBiomedicalEngineering 18, 26–49. doi:10.1109/RBME.2024.3449790

    work page doi:10.1109/rbme.2024.3449790 2025

  9. [9]

    Improve P300-speller performance by online tuning stimulus onset asynchrony (SOA)

    Gao, P., Huang, Y., He, F., Qi, H., 2021. Improve P300-speller performance by online tuning stimulus onset asynchrony (SOA). Journal of Neural Engineering 18. doi:10.1088/1741-2552/ac2f04

    work page doi:10.1088/1741-2552/ac2f04 2021

  10. [10]

    Enhancing SSVEP-BCI Performance Under Fatigue State Using Dynamic Stop- ping Strategy

    Han, Y., Ke, Y., Wang, R., Wang, T., Ming, D., 2024. Enhancing SSVEP-BCI Performance Under Fatigue State Using Dynamic Stop- ping Strategy. IEEE Transactions on Neural Systems and Rehabilita- tion Engineering 32, 1407–1415. doi:10.1109/TNSRE.2024.3380635

    work page doi:10.1109/tnsre.2024.3380635 2024

  11. [11]

    A Dynamic Recursive Unified Internet Design (DRUID),

    Hoffmann, U., Vesin, J.M., Ebrahimi, T., Diserens, K., 2008. An efficient P300-based brain–computer interface for disabled subjects. Journal of Neuroscience Methods 167, 115–125. doi:10.1016/j. jneumeth.2007.03.005

    work page doi:10.1016/j 2008

  12. [12]

    Two-dimensional auditory p300 speller with predictive text system, in: 2010 Annual International Conference of the IEEE Engineering in Medicine and Biology, pp

    Höhne, J., Schreuder, M., Blankertz, B., Tangermann, M., 2010. Two-dimensional auditory p300 speller with predictive text system, in: 2010 Annual International Conference of the IEEE Engineering in Medicine and Biology, pp. 4185–4188. doi:10.1109/IEMBS.2010. 5627379

    work page doi:10.1109/iembs.2010 2010

  13. [13]

    Jiménez, J., Rodríguez, F.B., In press. A Novel Approach to Char- acterizing and Analyzing the Speed-Accuracy Trade-off in Online P300-Based Brain-Computer Interfaces, in: Brain Informatics 2025, LectureNotesinComputerScience. doi:10.1007/978-981-95-9578-5_ 12

    work page doi:10.1007/978-981-95-9578-5_ 2025

  14. [14]

    Error-aware BCIs

    Kalaganis, F.P., Chatzilari, E.P., Laskaris, N.A., Nikolopoulos, S.A., Kompatsiaris, I.A., 2024. Error-aware BCIs. Signal Processing to Drive Human-Computer Interaction: EEG and eye-controlled inter- faces , 231–259doi:10.1049/PBCE129E_ch11

    work page doi:10.1049/pbce129e_ch11 2024

  15. [15]

    A well-conditioned estimator for large- dimensional covariance matrices

    Ledoit, O., Wolf, M., 2004. A well-conditioned estimator for large- dimensional covariance matrices. Journal of Multivariate Analysis 88, 365–411. doi:10.1016/S0047-259X(03)00096-4

    work page doi:10.1016/s0047-259x(03)00096-4 2004

  16. [16]

    A review of rapid serial visual presentation- based brain–computer interfaces

    Lees, S., Dayan, N., Cecotti, H., McCullagh, P., Maguire, L., Lotte, F., Coyle, D., 2018. A review of rapid serial visual presentation- based brain–computer interfaces. Journal of Neural Engineering 15, 021001. doi:10.1088/1741-2552/aa9817

    work page doi:10.1088/1741-2552/aa9817 2018

  17. [17]

    A review of classification algorithms for EEG-based brain–computer interfaces: A 10 year update

    Lotte, F., Bougrain, L., Cichocki, A., Clerc, M., Congedo, M., Rako- tomamonjy, A., Yger, F., 2018. A review of classification algorithms for EEG-based brain–computer interfaces: A 10 year update. Journal of Neural Engineering 15, 031005. doi:10.1088/1741-2552/aab2f2

    work page doi:10.1088/1741-2552/aab2f2 2018

  18. [18]

    Chapter 8

    Luck, S.J., 2014. Chapter 8. Baseline Correction, Averaging, and Time–Frequency Analysis, in: An Introduction to the Event-Related Potential Technique. 2 ed.. MIT Press, Cambridge, MA, USA, pp. 258–260

    2014

  19. [19]

    BCI-Utility Metric for Asynchronous P300 Brain-Computer Interface Systems

    Ma, G., Kang, J., Thompson, D.E., Huggins, J.E., 2023. BCI-Utility Metric for Asynchronous P300 Brain-Computer Interface Systems. IEEETransactionsonNeuralSystemsandRehabilitationEngineering 31, 3968–3977. doi:10.1109/TNSRE.2023.3322125

    work page doi:10.1109/tnsre.2023.3322125 2023

  20. [20]

    A general framework for brain-computer interfacedesign

    Mason, S., Birch, G., 2003. A general framework for brain-computer interfacedesign. IEEETransactionsonNeuralSystemsandRehabil- itation Engineering 11, 70–85. doi:10.1109/TNSRE.2003.810426

    work page doi:10.1109/tnsre.2003.810426 2003

  21. [21]

    Scikit-learn: Machine learning in Python

    Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., Duchesnay, E., 2011. Scikit-learn: Machine learning in Python. Journal of Machine Learning Research 12, 2825–2830

    2011

  22. [22]

    Clinical Neurophysiology 118, 2128–2148

    Polich,J.,2007.UpdatingP300:AnintegrativetheoryofP3aandP3b. Clinical Neurophysiology 118, 2128–2148. doi:10.1016/j.clinph. 2007.04.019

    work page doi:10.1016/j.clinph 2007

  23. [23]

    Evaluation of the performances of different P300 based brain–computer interfaces by means of the efficiency metric

    Quitadamo,L.R.,Abbafati,M.,Cardarilli,G.C.,Mattia,D.,Cincotti, F., Babiloni, F., Marciani, M.G., Bianchi, L., 2012. Evaluation of the performances of different P300 based brain–computer interfaces by means of the efficiency metric. Journal of Neuroscience Methods 203, 361–368. doi:10.1016/j.jneumeth.2011.10.010

    work page doi:10.1016/j.jneumeth.2011.10.010 2012

  24. [24]

    Brain computer interface: Control signals review

    Ramadan, R.A., Vasilakos, A.V., 2017. Brain computer interface: Control signals review. Neurocomputing 223, 26–44. doi:10.1016/ j.neucom.2016.10.024

    2017

  25. [25]

    (Eds.), 2020

    Ramsey, N.F., Millán, J. (Eds.), 2020. Brain-Computer Interfaces. volume 168 ofHandbook of Clinical Neurology. 1 ed., Elsevier, San Diego

    2020

  26. [26]

    Optimizing event-related potential based brain-computer interfaces: A systematic evaluation of dynamic stop- ping methods

    Schreuder, M., Höhne, J., Blankertz, B., Haufe, S., Dickhaus, T., Tangermann, M., 2013. Optimizing event-related potential based brain-computer interfaces: A systematic evaluation of dynamic stop- ping methods. Journal of Neural Engineering 10, 036025. doi:10. 1088/1741-2560/10/3/036025

    2013

  27. [27]

    NeuroImage289, 120548

    Shi, N., Miao, Y., Huang, C., Li, X., Song, Y., Chen, X., Wang, Y., Gao,X.,2024.Estimatingandapproachingthemaximuminformation rateofnoninvasivevisualbrain-computerinterface. NeuroImage289, 120548. doi:10.1016/j.neuroimage.2024.120548

    work page doi:10.1016/j.neuroimage.2024.120548 2024

  28. [28]

    Soetedjo,J.,Arnold,C.,Pouratian,N.,Speier,W.,Keluo-Udeke,O.S.,

  29. [29]

    Maintaining High Accuracy General P300 Speller Using the Language Modeling and Dynamic Stopping, in: 2020 IEEE 20th In- ternationalConferenceonBioinformaticsandBioengineering(BIBE 2020), IEEE, New York. pp. 365–368. doi:10.1109/BIBE50027.2020. 00066

    work page doi:10.1109/bibe50027.2020 2020

  30. [30]

    An Asyn- chronous Training-Free SSVEP-BCI Detection Algorithm for Non- Equal Prior Probability Scenarios

    Wang, J., Cui, Y., Zhang, H., Wu, H., Yang, C., 2024a. An Asyn- chronous Training-Free SSVEP-BCI Detection Algorithm for Non- Equal Prior Probability Scenarios. IEEE Transactions on Neural SystemsandRehabilitationEngineering32,4120–4130. doi:10.1109/ TNSRE.2024.3496727

    work page arXiv 2024

  31. [31]

    Average TimeConsumptionperCharacter:APracticalPerformanceMetricfor Generic Synchronous BCI Spellers

    Wang, Z., Hu, H., Zhou, T., Xu, T., Zhao, X., 2024b. Average TimeConsumptionperCharacter:APracticalPerformanceMetricfor Generic Synchronous BCI Spellers. IEEE Transactions on Biomedi- cal Engineering 71, 2684–2698. doi:10.1109/TBME.2024.3387469

    work page doi:10.1109/tbme.2024.3387469 2024

  32. [32]

    Brain–computerinterfaces for communication and control

    Wolpaw, J.R., Birbaumer, N., McFarland, D.J., Pfurtscheller, G., Vaughan, T.M.,2002. Brain–computerinterfaces for communication and control. Clinical Neurophysiology 113, 767–791. doi:10.1016/ S1388-2457(02)00057-3

    2002

  33. [33]

    EEGDatasetforRSVP and P300 Speller Brain-Computer Interfaces

    Won,K.,Kwon,M.,Ahn,M.,Jun,S.C.,2022. EEGDatasetforRSVP and P300 Speller Brain-Computer Interfaces. Scientific Data 9, 388. doi:10.1038/s41597-022-01509-w

    work page doi:10.1038/s41597-022-01509-w 2022

  34. [34]

    Electroencephalogram based brain- computerinterface:Applications,challenges,andopportunities.Mul- timedia Tools and Applications 82, 47003–47047

    Yadav, H., Maini, S., 2023. Electroencephalogram based brain- computerinterface:Applications,challenges,andopportunities.Mul- timedia Tools and Applications 82, 47003–47047. doi:10.1007/ s11042-023-15653-x

    2023

  35. [35]

    A study of the existing problems of estimating the information transfer rate in online brain-computer interfaces

    Yuan, P., Gao, X., Allison, B., Wang, Y., Bin, G., Gao, S., 2013. A study of the existing problems of estimating the information transfer rate in online brain-computer interfaces. Journal of Neural Engineer- ing 10, 026014. doi:10.1088/1741-2560/10/2/026014. J. Jiménez et al.:Preprint submitted to ElsevierPage 13 of 13

    work page doi:10.1088/1741-2560/10/2/026014 2013