Brain–computer interfaces (BCIs) translate brain activity into commands, enabling applications in communication, control, and neurorehabilitation. A major challenge in noninvasive BCIs is balancing classification performance with interpretability, as many approaches prioritize accuracy while overlooking the neural mechanisms underlying their predictions.
Methods:
In this study, we conduct a meta-analysis of feature interpretability across widely used methods in motor imagery (MI)-based BCIs, including power spectral density, common spatial patterns (CSP), Riemannian geometry, and functional connectivity. Specifically, we explore how network topology and spatial organization contribute to MI decoding by investigating brain network lateralization.
Results:
Through evaluations on multiple EEG-based BCI datasets, our results confirm the superior classification performance of CSP and Riemannian methods. However, network lateralization provides stronger neurophysiological plausibility, revealing robust lateralization patterns in sensorimotor and frontal regions contralateral to imagined movements.
Discussion:
These findings underscore the potential of connectivity-based features as a complementary tool for enhancing interpretability, supporting the development of more transparent and clinically relevant MI-based BCIs.
Impact Statement
This study addresses a critical gap in motor imagery-based brain–computer interfaces (BCIs) by systematically evaluating and comparing the interpretability of widely used methods, including power spectral density, common spatial pattern, Riemannian geometry, and functional connectivity. By analyzing these approaches across wide-ranging datasets, we offer valuable insights into the underlying neural mechanisms driving their performance. Our findings contribute to enhancing the transparency and biological relevance of BCI systems, ultimately advancing the development of more clinically meaningful and neurophysiologically interpretable BCIs.
AristimunhaB, CarraraI, GuetschelP, et al.Mother of all BCI benchmarks. J Neural Eng, 2025; doi: 10.5281/zenodo.10034223
2.
BartolomeoP, MalkinsonTS. Hemispheric lateralization of attention processes in the human brain. Curr Opin Psychol, 2019; 29:90–96.
3.
BarzegaranE, KnyazevaMG. Functional connectivity analysis in EEG source space: The choice of method. PLoS One, 2017; 12(7):e0181105; doi: 10.1371/journal.pone.0181105
4.
BashashatiA, FatourechiM, WardRK, et al.A survey of signal processing algorithms in brain-computer interfaces based on electrical brain signals. J Neural Eng, 2007; 4(2):R32–R57; doi: 10.1088/1741-2560/4/2/R03
5.
BastosAM, SchoffelenJ-M. A tutorial review of functional connectivity analysis methods and their interpretational pitfalls. Front Syst Neurosci, 2015; 9:175; doi: 10.3389/fnsys.2015.00175
6.
BiasiucciA, LeebR, IturrateI, et al.Brain-actuated functional electrical stimulation elicits lasting arm motor recovery after stroke. Nat Commun, 2018; 9(1):2421; doi: 10.1038/s41467-018-04673-z
7.
BarachantA, BonnetS. Channel Selection Procedure Using Riemannian Distance for BCI Applications. In: 2011 5th International IEEE/EMBS Conference on Neural Engineering, 2011; pp. 348–351; doi: 10.1109/NER.2011.5910558
8.
BarachantA, BonnetS, CongedoM, et al.Classification of covariance matrices using a Riemannian-based kernel for BCI applications. Neurocomputing (Amst), 2013; 112:172–178; doi: 10.1016/j.neucom.2012.12.039
9.
BarachantA, BonnetS, CongedoM, et al.Common Spatial Pattern Revisited by Riemannian Geometry. In: 2010 IEEE International Workshop on Multimedia Signal Processing, 2010a; pp. 472–476; doi: 10.1109/MMSP.2010.5662067
10.
BarachantA, BonnetS, CongedoM, et al.Multiclass brain–computer interface classification by Riemannian geometry. IEEE Trans Biomed Eng, 2012; 59(4):920–928; doi: 10.1109/TBME.2011.2172210
11.
BarachantA, BonnetS, CongedoM, et al.Riemannian Geometry Applied to BCI Classification. In: Latent Variable Analysis and Signal Separation. (VigneronV, ZarzosoV, MoreauE eds). Lecture Notes in Computer Science Springer: Berlin, Heidelberg; 2010b; pp. 629–636; doi: 10.1007/978-3-642-15995-4_78
12.
BoccalettiS, BianconiG, CriadoR, et al.The structure and dynamics of multilayer networks. Phys Rep, 2014; 544(1):1–122; doi: 10.1016/j.physrep.2014.07.001
13.
BrunnerC, SchererR, GraimannB, et al.Online control of a brain-computer interface using phase synchronization. IEEE Trans Biomed Eng, 2006; 53(12 Pt 1):2501–2506; doi: 10.1109/TBME.2006.881775
14.
CaoL, WangW, HuangC, et al.An effective fusing approach by combining connectivity network pattern and temporal-spatial analysis for EEG-based BCI rehabilitation. IEEE Trans Neural Syst Rehabil Eng, 2022; 30:2264–2274; doi: 10.1109/TNSRE.2022.3198434
15.
CarlsonT, del R. MillanJ. Brain-controlled wheelchairs: A robotic architecture. IEEE Robot Automat Mag, 2013; 20(1):65–73; doi: 10.1109/MRA.2012.2229936
16.
CarterGC. Coherence and time delay estimation. Proc IEEE, 1987; 75(2):236–255; doi: 10.1109/PROC.1987.13723
17.
CattaiT, ColonneseS, CorsiM-C, et al.Phase/amplitude synchronization of brain signals during motor imagery BCI tasks. IEEE Trans Neural Syst Rehabil Eng, 2021; 29:1168–1177; doi: 10.1109/TNSRE.2021.3088637
18.
ChoH, AhnM, AhnS, et al.EEG datasets for motor imagery brain-computer interface. Gigascience, 2017; 6(7):1–8; doi: 10.1093/gigascience/gix034
19.
ChuY, ZhaoX, ZouY, et al.Decoding multiclass motor imagery EEG from the same upper limb by combining Riemannian geometry features and partial least squares regression. J Neural Eng, 2020; 17(4):46029; doi: 10.1088/1741-2552/aba7cd
20.
BlankertzB, CurioG, MüllerK-R. Classifying Single Trial EEG: Towards Brain Computer Interfacing. 2002.
21.
BlankertzB, DornhegeG, KrauledatM, et al.The Berlin Brain-Computer Interface: Report from the Feedback Sessions. 2005.
22.
BlankertzB, TomiokaR, LemmS, et al.Optimizing spatial filters for robust EEG single-trial analysis. IEEE Signal Process Mag, 2008; 25(1):41–56; doi: 10.1109/MSP.2008.4408441
23.
CorsiM-C, ChevallierS, De Vico FallaniF, et al.Functional Connectivity Ensemble Method to Enhance BCI Performance (FUCONE). IEEE Trans Biomed Eng, 2022; 69(9):2826–2838; doi: 10.1109/TBME.2022.3154885
24.
CurtisCE, D’EspositoM. Persistent activity in the prefrontal cortex during working memory. Trends Cogn Sci, 2003; 7(9):415–423; doi: 10.1016/s1364-6613(03)00197-9
25.
CongedoM, BarachantA, BhatiaR. Riemannian geometry for EEG-based brain-computer interfaces; a primer and a review. Brain-Computer Interfaces, 2017a;4(3):155–174; doi: 10.1080/2326263X.2017.1297192
26.
CongedoM, Coelho RodriguesPL, BouchardF, et al.A Closed-Form Unsupervised Geometry-Aware Dimensionality Reduction Method in the Riemannian Manifold of SPD Matrices. In: EMBC 2017 - 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society IEEE: Jeju Island, South Korea; 2017b; pp. 3198–3201.
27.
De Vico FallaniF, RichiardiJ, ChavezM, et al.Graph analysis of functional brain networks: Practical issues in translational neuroscience. Philos Trans R Soc Lond B Biol Sci, 2014; 369(1653):20130521; doi: 10.1098/rstb.2013.0521
28.
DehaeneS, KerszbergM, ChangeuxJ-P. A neuronal model of a global workspace in effortful cognitive tasks. Proc Natl Acad Sci U S A, 1998; 95(24):14529–14534; doi: 10.1073/pnas.95.24.14529
29.
DesboisA, VenotT, De Vico FallaniF, et al.HappyFeat—An interactive and efficient BCI framework for clinical applications. Softw Impacts, 2024; 19:100610.
30.
EdelmanBJ, BaxterB, HeB. EEG source imaging enhances the decoding of complex right-hand motor imagery tasks. IEEE Trans Biomed Eng, 2016; 63(1):4–14; doi: 10.1109/TBME.2015.2467312
31.
FarrerC, ValentinG, HupéJM. The time windows of the sense of agency. Conscious Cogn, 2013; 22(4):1431–1441; doi: 10.1016/j.concog.2013.09.010
32.
FeinG, RazJ, BrownFFet al.Common reference coherence data are confounded by power and phase effects. Electroencephalography and Clinical Neurophysiology, 1988; 69(6):581–584; doi: 10.1016/0013-4694(88)90171-x
33.
FengZ, QianL, HuH, et al.Functional Connectivity for Motor Imaginary Recognition in Brain-Computer Interface. In: 2020 IEEE International Conference on Systems, Man, and Cybernetics (SMC) 2020; pp. 3678–3682; doi: 10.1109/SMC42975.2020.9283075
34.
GaoQ, DuanX, ChenH. Evaluation of effective connectivity of motor areas during motor imagery and execution using conditional Granger causality. Neuroimage, 2011; 54(2):1280–1288; doi: 10.1016/j.neuroimage.2010.08.071
35.
García-PrietoJ, BajoR, PeredaE. Efficient computation of functional brain networks: Toward real-time functional connectivity. Front Neuroinform, 2017; 11:8.
36.
GaumeA, DreyfusG, VialatteF-B. A cognitive brain–computer interface monitoring sustained attentional variations during a continuous task. Cogn Neurodyn, 2019; 13(3):257–269; doi: 10.1007/s11571-019-09521-4
37.
GazzanigaMS. Cerebral specialization and interhemispheric communication: Does the corpus callosum enable the human condition? Brain, 2000; 123 (Pt 7):1293–1326; doi: 10.1093/brain/123.7.1293
38.
GoldbergerAL, AmaralLA, GlassL, et al.PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource for complex physiologic signals. Circulation, 2000; 101(23):E215–E220; doi: 10.1161/01.cir.101.23.e215
39.
Gonzalez-AstudilloJ, CattaiT, BassignanaG, et al.Network-based brain–computer interfaces: Principles and applications. J Neural Eng, 2021; 18(1):011001; doi: 10.1088/1741-2552/abc760
40.
GottsSJ, JoHJ, WallaceGL, et al.Two distinct forms of functional lateralization in the human brain. Proc Natl Acad Sci U S A, 2013; 110(36); doi: 10.1073/pnas.1302581110
41.
GrèzesJ, DecetyJ. Functional anatomy of execution, mental simulation, observation, and verb generation of actions: A meta-analysis. Hum Brain Mapp, 2001; 12(1):1–19; doi: 10.1002/1097-0193(200101)12:1<1::aid-hbm10>3.0.co;2-v
42.
Grosse-WentrupM, LiefholdC, GramannK, et al.Beamforming in noninvasive brain–computer interfaces. IEEE Trans Biomed Eng, 2009; 56(4):1209–1219; doi: 10.1109/TBME.2008.2009768
43.
GuillotA, ColletC, NguyenVA, et al.Brain activity during visual versus kinesthetic imagery: An fMRI study. Hum Brain Mapp, 2009; 30(7):2157–2172; doi: 10.1002/hbm.20658
44.
HamediM, SallehS-H, NoorAM. Electroencephalographic motor imagery brain connectivity analysis for BCI: A review. Neural Comput, 2016; 28(6):999–1041; doi: 10.1162/NECO_a_00838
45.
HétuS, GrégoireM, SaimpontA, et al.The neural network of motor imagery: An ALE meta-analysis. Neurosci Biobehav Rev, 2013; 37(5):930–949; doi: 10.1016/j.neubiorev.2013.03.017
46.
JatoiMA, KamelN, MalikAS, et al.A survey of methods used for source localization using EEG signals. Biomed Signal Process Control, 2014; 11:42–52; doi: 10.1016/j.bspc.2014.01.009
JeannerodM, FrakV. Mental imaging of motor activity in humans. Curr Opin Neurobiol, 1999; 9(6):735–739; doi: 10.1016/S0959-4388(99)00038-0
49.
JeannerodM. Mental imagery in the motor context. Neuropsychologia, 1995; 33(11):1419–1432; doi: 10.1016/0028-3932(95)00073-c
50.
KarolisVR, CorbettaM, Thiebaut de SchottenM. The architecture of functional lateralisation and its relationship to callosal connectivity in the human brain. Nat Commun, 2019; 10(1):1417; doi: 10.1038/s41467-019-09344-1
51.
KimMS, ParkH, KwonI, et al.Efficacy of brain-computer interface training with motor imagery-contingent feedback in improving upper limb function and neuroplasticity among persons with chronic stroke: A double-blinded, parallel-group, randomized controlled trial. J Neuroeng Rehabil, 2025; 22(1):1; doi: 10.1186/s12984-024-01535-2
52.
KruseA, SuicaZ, TaeymansJ, et al.Effect of brain-computer interface training based on non-invasive electroencephalography using motor imagery on functional recovery after stroke—A systematic review and meta-analysis. BMC Neurol, 2020; 20(1):385; doi: 10.1186/s12883-020-01960-5
53.
LarzabalC, AuboirouxV, KarakasS, et al.The Riemannian spatial pattern method: Mapping and clustering movement imagery using Riemannian geometry. J Neural Eng, 2021; 18(5):056014; doi: 10.1088/1741-2552/abf291
54.
LawhernVJ, SolonAJ, WaytowichNR, et al.EEGNet: A compact convolutional neural network for EEG-based brain–computer interfaces. J Neural Eng, 2018; 15(5):56013; doi: 10.1088/1741-2552/aace8c
55.
LeeM-H, KwonO-Y, KimY-J, et al.EEG dataset and OpenBMI toolbox for three BCI paradigms: An investigation into BCI illiteracy. Gigascience, 2019; 8(5):giz002.
56.
LeeS-H, LeeM, LeeS-W. Neural decoding of imagined speech and visual imagery as intuitive paradigms for BCI communication. IEEE Trans Neural Syst Rehabil Eng, 2020; 28(12):2647–2659; doi: 10.1109/TNSRE.2020.3040289
57.
LeeuwisN, YoonS, AlimardaniM. Functional connectivity analysis in motor-imagery brain computer interfaces. Front Hum Neurosci, 2021; 15:732946; doi: 10.3389/fnhum.2021.732946
58.
LiJ, WangY, ZhangL, et al.Decoding EEG in cognitive tasks with time-frequency and connectivity masks. IEEE Trans Cogn Dev Syst, 2016; 8(4):298–308; doi: 10.1109/TCDS.2016.2555952
59.
LiuH, StufflebeamSM, SepulcreJ, et al.Evidence from intrinsic activity that asymmetry of the human brain is controlled by multiple factors. Proc Natl Acad Sci U S A, 2009; 106(48):20499–20503; doi: 10.1073/pnas.0908073106
60.
LotteF, CongedoM, LécuyerA, et al.A review of classification algorithms for EEG-based brain-computer interfaces. J Neural Eng, 2007; 4(2):R1–R13; doi: 10.1088/1741-2560/4/2/R01
61.
LotteF. A Tutorial on EEG Signal-Processing Techniques for Mental-State Recognition in Brain–Computer Interfaces. In: Guide to Brain-Computer Music Interfacing. (MirandaER and CastetJ. eds) Springer London: London; 2014; pp. 133–161; doi: 10.1007/978-1-4471-6584-2_7
ManeR, ChewE, ChuaK, et al.FBCNet: A multi-view convolutional neural network for brain-computer interface. arXiv Preprint arXiv, 2021; doi: 10.48550/arXiv.2104.01233
64.
ManeR, ChouhanT, GuanC. BCI for stroke rehabilitation: Motor and beyond. J Neural Eng, 2020; 17(4):41001; doi: 10.1088/1741-2552/aba162
65.
MokienkoO, ChervyakovA, KulikovaS, et al.Increased motor cortex excitability during motor imagery in brain-computer interface trained subjects. Front Comput Neurosci, 2013; 7:168.
66.
MonacoS, MalfattiG, CulhamJC, et al.Decoding motor imagery and action planning in the early visual cortex: Overlapping but distinct neural mechanisms. Neuroimage, 2020; 218:116981; doi: 10.1016/j.neuroimage.2020.116981
67.
NeuperC, PfurtschellerG. Event-related dynamics of cortical rhythms: Frequency-specific features and functional correlates. Int J Psychophysiol, 2001; 43(1):41–58; doi: 10.1016/s0167-8760(01)00178-7
68.
Patrick-KruegerKM, BurkhartI, Contreras-VidalJL. The state of clinical trials of implantable brain–computer interfaces. Nat Rev Bioeng, 2024; 3(1):50–67; doi: 10.1038/s44222-024-00239-5
69.
PfurtschellerG, AranibarA. Event-related cortical desynchronization detected by power measurements of scalp EEG. Electroencephalogr Clin Neurophysiol, 1977; 42(6):817–826; doi: 10.1016/0013-4694(77)90235-8
70.
PfurtschellerG, Lopes da SilvaFH. Event-related EEG/MEG synchronization and desynchronization: Basic principles. Clin Neurophysiol, 1999; 110(11):1842–1857; doi: 10.1016/s1388-2457(99)00141-8
71.
PfurtschellerG, NeuperC. Motor imagery and direct brain-computer communication. Proc IEEE, 2001; 89(7):1123–1134; doi: 10.1109/5.939829
72.
PichiorriF, MoroneG, PettiM, et al.Brain–computer interface boosts motor imagery practice during stroke recovery. Ann Neurol, 2015; 77(5):851–865; doi: 10.1002/ana.24390
73.
QuT, JinJ, XuR, et al.Riemannian distance based channel selection and feature extraction combining discriminative time-frequency bands and Riemannian tangent space for MI-BCIs. J Neural Eng, 2022; 19(5):056025; doi: 10.1088/1741-2552/ac9338
74.
RobinsonN, ManeR, ChouhanT, et al.Emerging trends in BCI-robotics for motor control and rehabilitation. Curr Opin Biomed Eng, 2021; 20:100354; doi: 10.1016/j.cobme.2021.100354
75.
RodriguesP, BouchardF, CongedoM, et al.Dimensionality reduction for BCI classification using Riemannian geometry. Intr Brain-Computer Interface Confer, 2017; doi: 10.3217/978-3-85125-533-1-16
76.
RuessM, ThomaschkeR, KieselA. The time course of intentional binding. Atten Percept Psychophys, 2017; 79(4):1123–1131; doi: 10.3758/s13414-017-1292-y
77.
SabatéM, GonzálezB, RodríguezM. Brain lateralization of motor imagery: Motor planning asymmetry as a cause of movement lateralization. Neuropsychologia, 2004; 42(8):1041–1049; doi: 10.1016/j.neuropsychologia.2003.12.015
78.
SchirrmeisterRT, SpringenbergJT, FiedererLDJ, et al.Deep learning with convolutional neural networks for EEG decoding and visualization. Hum Brain Mapp, 2017; 38(11):5391–5420; doi: 10.1002/hbm.23730
79.
ShinJ, von LühmannA, BlankertzB, et al.Open access dataset for EEG+ NIRS single-trial classification. IEEE Trans Neural Syst Rehabil Eng, 2017; 25(10):1735–1745.
80.
SlepianD. Prolate spheroidal wave functions, Fourier analysis, and uncertainty—V: The discrete case. Bell Syst Tech J, 1978; 57(5):1371–1430; doi: 10.1002/j.1538-7305.1978.tb02104.x
81.
SmithEE, ReznikSJ, StewartJL, et al.Assessing and conceptualizing Frontal EEG asymmetry: An updated primer on recording, processing, analyzing, and interpreting frontal alpha asymmetry. Int J Psychophysiol, 2017; 111:98–114.
82.
SongY, ZhengQ, LiuB, et al.EEG conformer: Convolutional transformer for EEG decoding and visualization. IEEE Trans Neural Syst Rehabil Eng, 2023; 31:710–719; doi: 10.1109/TNSRE.2022.3230250
83.
TangermannM, MüllerK-R, AertsenA, et al.Review of the BCI competition IV. Front Neurosci, 2012; 6:55; doi: 10.3389/fnins.2012.00055
84.
VanderahT, GouldDJ. Nolte’s The Human Brain E-Book: An Introduction to Its Functional Anatomy. Elsevier Health Sciences; 2015.
85.
VansteenselMJ, PelsEGM, BleichnerMG, et al.Fully implanted brain-computer interface in a locked-in patient with ALS. N Engl J Med, 2016; 375(21):2060–2066; doi: 10.1056/NEJMoa1608085
86.
VenotT, DesboisA, and VicoFallaniF. Dynamic Brain Networks in Motor Imagery-Based BCI. Graz BCI Conference Verlag der Technischen Universität Graz; 2024.
87.
VidalJJ. Toward direct brain-computer communication. Annu Rev Biophys Bioeng, 1973; 2:157–180; doi: 10.1146/annurev.bb.02.060173.001105
88.
WangT, MaoR, LiuS, et al.Explainable multi-frequency and multi-region fusion model for affective brain-computer interfaces. Information Fusion, 2025; 118:102971; doi: 10.1016/j.inffus.2025.102971
89.
WangZ-M, ChenZ-Y, ZhangJ. EEG emotion recognition based on PLV-rich-club dynamic brain function network. Appl Intell, 2023; 53(14):17327–17345; doi: 10.1007/s10489-022-04366-7
90.
WolbersT, WeillerC, BüchelC. Contralateral coding of imagined body parts in the superior parietal lobe. Cereb Cortex, 2003; 13(4):392–399; doi: 10.1093/cercor/13.4.392
91.
WolpawJR, BirbaumerN, McFarlandDJ, et al.Brain-computer interfaces for communication and control. Clin Neurophysiol, 2002; 113(6):767–791; doi: 10.1016/s1388-2457(02)00057-3
92.
XuJ, Grosse-WentrupM, JayaramV. Tangent space spatial filters for interpretable and efficient Riemannian classification. J Neural Eng, 2020; 17(2):26043; doi: 10.1088/1741-2552/ab839e
93.
XuL, ZhangH, HuiM, et al.Motor execution and motor imagery: A comparison of functional connectivity patterns based on graph theory. Neuroscience, 2014; 261:184–194; doi: 10.1016/j.neuroscience.2013.12.005
94.
YgerF, BerarM, LotteF. Riemannian approaches in brain-computer interfaces: A review. IEEE Trans Neural Syst Rehabil Eng, 2017; 25(10):1753–1762; doi: 10.1109/TNSRE.2016.2627016
95.
YiW, QiuS, WangK, et al.Evaluation of EEG oscillatory patterns and cognitive process during simple and compound limb motor imagery. PLoS One, 2014; 9(12):e114853.
96.
ZapałaD, Zabielska-MendykE, AugustynowiczP, et al.The effects of handedness on sensorimotor rhythm desynchronization and motor-imagery BCI control. Sci Rep, 2020; 10(1):2087; doi: 10.1038/s41598-020-59222-w
97.
ZhanG, ChenS, JiY, et al.EEG-based brain network analysis of chronic stroke patients after BCI rehabilitation training. Front Hum Neurosci, 2022; 16:909610.
98.
ZhangH, ChavarriagaR, GoelMK, et al.Improved Recognition of Error Related Potentials Through the Use of Brain Connectivity Features. In: 2012 Annual International Conference of the IEEE Engineering in Medicine and Biology Society IEEE; 2012; pp. 6740–6743.
99.
ZhouB, WuX, LvZ, et al.A fully automated trial selection method for optimization of motor imagery based brain-computer interface. PLoS One, 2016; 11(9):e0162657.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
