Electroencephalograph (EEG) based classification of mental arithmetic using explainable machine learning

• Autorzy: Aslam Murtaza , Rajbdad Fozia , Azmat Shoaib , Perveen Kausar , Naraghi-Pour Morteza , Xu Jian

Identyfikatory

DOI

10.1016/j.bbe.2025.02.002

• Języki publikacji: EN

Abstrakty

EN

Mental arithmetic can be helpful for the evaluation of neurodevelopmental disorders arising from atypical development of the brain. We propose a novel explainable machine learning method for classifying mental arithmetic calculation tasks from resting brain states and good from bad calculations using Electroencephalography. Empirical mode decomposition features are extracted from intrinsic mode functions of the average signals of all trials. Most relevant features to the mental arithmetic tasks are ranked by a random forest-based recursive feature elimination method. These features identify the changes in frequency bands of the brain rhythms, such as delta, theta, and alpha, during mental tasks for the first time in literature. These unique explainable features are also used to identify brain areas such as frontal, temporal, and occipital lobes involved in mental arithmetic tasks. Moreover, our approach describes the memory regions and that bad calculations excite the brain areas, mostly related to emotions such as frustration and anxiety due to stressful mental arithmetic. Using a random forest classifier, beating the state-of-the-art, this method achieved classification accuracies of 99.30 % and 98.33 % for resting vs calculation and good vs bad calculation brain tasks, respectively. Also, our method outperformed the state of art in handling the inter-subject variability and achieved 98.17 ± 0.47 % and 97.19 ± 0.95 % classification accuracies for resting vs calculation and good vs bad calculation tasks, respectively.

Słowa kluczowe

EN

electroencephalography; mental arithmetic; recursive feature elimination; empirical mode decomposition; explainable machine learning

PL

elektroencefalografia; selekcja cech; rozkład trybu empirycznego

• Wydawca: Nałęcz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences ; Elsevier
• Czasopismo: Biocybernetics and Biomedical Engineering
• Rocznik: 2025
• Tom: Vol. 45, iss. 2
• Strony: 154--169
• Opis fizyczny: Bibliogr. 71 poz., rys., tab., wykr.

Twórcy

autor

Aslam Murtaza

• Division of Electrical and Computer Engineering, Louisiana State University, Baton Rouge, USA

autor

Rajbdad Fozia

• Division of Electrical and Computer Engineering, Louisiana State University, Baton Rouge, USA

autor

Azmat Shoaib

• Department of Electrical and Computer Engineering, COMSATS University Islamabad, Islamabad, Pakistan

• https://orcid.org/0000-0002-0954-8334

autor

Perveen Kausar

• Department of Internal Medicine and Neuroscience, National Medical Commission (NMC), Royal Hospital, Sharjah, United Arab Emirates

autor

Naraghi-Pour Morteza

• Division of Electrical and Computer Engineering, Louisiana State University, Baton Rouge, USA

• https://orcid.org/0000-0003-1300-8733

autor

Xu Jian

• Division of Electrical and Computer Engineering, Louisiana State University, Baton Rouge, USA

Bibliografia

• [1] Liu S, Wang L, Gao RX. Cognitive neuroscience and robotics: Advancements and future research directions. Rob Comput Integr Manuf 2024;85:102610.
• [2] Núñez-Peña MI, Campos-Rodríguez C. Response monitoring in math-anxious individuals in an arithmetic task. Biol Psychol 2024;186:108759.
• [3] Boxum M, Voetterl H, van Dijk H, Gordon E, DeBeus R, Arnold LE, et al. Challenging the diagnostic value of theta/beta ratio: insights from an EEG subtyping meta-analytical approach in ADHD. Appl Psychophysiol Biofeedback 2024: 1–12.
• [4] Ortuño-Miró S, Molina-Rodríguez S, Belmonte C, Ibañez-Ballesteros J. Identifying ADHD boys by very-low frequency prefrontal fNIRS fluctuations during a rhythmic mental arithmetic task. J Neural Eng 2023;20(3):036018.
• [5] Polo-Blanco I, Suárez-Pinilla P, Goñi-Cervera J, Suárez-Pinilla M, Payá B. Comparison of mathematics problem-solving abilities in autistic and non-autistic children: The influence of cognitive profile. J Autism Dev Disord 2024;54(1): 353-65.
• [6] Bonifacci P, Serena B, Codeluppi F, Peri B. A longitudinal study of math skills in heritage bilingual children: profiles of strengths and weaknesses. Psychol Res 2025; 89(1):8.
• [7] Otálora Y, Taborda-Osorio H. Contributions of numeral knowledge, spatial reasoning, and numerical mapping to first graders’ arithmetic and geometric skills. Cogn Dev 2025;73:101520.
• [8] Träff U, Östergren R, Skagerlund K, Skagenholt M. Mental arithmetic skill development in primary school: The importance of number processing abilities and general cognitive abilities. J Exp Child Psychol 2025;252:106155.
• [9] Harrison AG, Beal AL, Armstrong IT, Gallagher A. Measurement of Working Memory on the Wechsler Adult Intelligence Scale: should We Subtract Arithmetic? Psychol Injury Law 2024;17(1):55-65.
• [10] Radević L, Milovanović I. Current trends in math anxiety research: a bibliometric approach. Int J Sci Math Educ 2024;22(6):1345-62.
• [11] Ren X, Libertus ME. Identifying the neural bases of math competence based on structural and functional properties of the human brain. J Cogn Neurosci 2023;35(8):1212-28.
• [12] Saini M, Satija U. State-of-the-art mental tasks classification based on electroencephalograms: a review. Physiol Meas 2023;44(6):06TR01.
• [13] Bergil N, Aydemir E, Barua PD, Baygin M, Dogan S, Tuncer T, et al. Automated mental arithmetic performance detection using quantum pattern-and triangle pooling techniques with EEG signals. Expert Syst Appl 2023;227:120306.
• [14] Sharma LD, Chhabra H, Chauhan U, Saraswat RK, Sunkaria RK. Mental arithmetic task load recognition using EEG signal and Bayesian optimized K-nearest neighbor. Int J Inf Technol 2021;13(6):2363-9.
• [15] Saini M, Satija U, Upadhayay MD. Discriminatory features based on wavelet energy for effective analysis of electroencephalogram during mental tasks. Circuit Syst Signal Proc 2022;41(10):5827-55.
• [16] Rajendran VG, Jayalalitha S, Adalarasu K, Mathi R. Machine learning based human mental state classification using wavelet packet decomposition-an EEG study. Multimed Tools Appl 2024:1-20.
• [17] Bonança GM, Gerhardt GJ, Molan AL, Oliveira LM, Jarola GM, Schönwald SV, et al. EEG alpha and theta time-frequency structure during a written mathematical task. Med Biol Eng Compu 2024;62(6):1869-85.
• [18] Youssofzadeh V, Roy S, Chowdhury A, Izadysadr A, Parkkonen L, Raghavan M, et al. Mapping and decoding cortical engagement during motor imagery, mental arithmetic, and silent word generation using MEG. Hum Brain Mapp 2023;44(8): 3324-42.
• [19] Karnan H, Uma Maheswari D, Priyadharshini D, Laushya S, Thivyaprakas TK. Cognizance detection during mental arithmetic task using statistical approach. Comput Methods Biomech Biomed Eng 2023:1-14.
• [20] Al-jumaili S. Efficient mental arithmetic classification using approximate entropy features and machine learning classifiers. Aurum J Health Sci 2024;5(3):109-20.
• [21] Ergün E, Aydemir O. A new evolutionary preprocessing approach for classification of mental arithmetic based EEG signals. Cogn Neurodyn 2020;14:609-17.
• [22] Kumar S, Shekhar S, Agarwal P. Cognitive capability identification in performing mental tasks using EEG-based coherence. Int J Syst Assur Eng Manage 2023;14(1): 334-42.
• [23] Behrouzi T, Hatzinakos D. Graph variational auto-encoder for deriving EEG-based graph embedding. Pattern Recogn 2022;121:108202.
• [24] Gangapuram H, Manian V. Electroencephalogram functional connectivity analysis and classification of mental arithmetic working memory task. Signals 2024;5(2): 296-325.
• [25] Yin Z, Tang J, Zhang J, Wang Z, Zhang J, Wang Y, et al. Generic Mental Workload Measurement Using a Shared Spatial Map Network with Different EEG Channel Layouts. IEEE Trans Instrum Meas 2024.
• [26] Yedukondalu J, Sharma LD. Cognitive load detection using circulant singular spectrum analysis and Binary Harris Hawks Optimization based feature selection. Biomed Signal Process Control 2023;79:104006.
• [27] Shah, D., Gopan K, G., & Sinha, N. (2022). An investigation of the multidimensional (1D vs. 2D vs. 3D) analyses of EEG signals using traditional methods and deep learning-based methods. Frontiers in Signal Processing, 2, 936790.
• [28] Fatimah B, Javali A, Ansar H, Harshitha BG, Kumar H. In: July). Mental arithmetic task classification using fourier decomposition method. IEEE; 2020. p. 0046-50.
• [29] Coman DA, Ionita S, Lita I. Evaluation of EEG Signals by Spectral Peak Methods and Statistical Correlation for Mental State Discrimination Induced by Arithmetic Tasks. Sensors 2024;24(11):3316.
• [30] Bergil E, Oral C, Ergül EU. Classification of arithmetic mental task performances using EEG and ECG signals. J Supercomput 2023;79(14):15535-47.
• [31] Babu, T. A., Gadde, S., Ravi, S., Rao, K. V., Mamillu, Y., & Krishna, D. (2022, March). Analysis of mental task ability in students based on electroencephalography signals. In 2022 IEEE International Conference on Signal Processing, Informatics, Communication and Energy Systems (SPICES) (Vol. 1, pp. 274-278). IEEE.
• [32] Yedukondalu J, Sunkara K, Radhika V, Kondaveeti S, Anumothu M, Murali Krishna Y. Cognitive load detection through EEG lead wise feature optimization and ensemble classification. Sci Rep 2025;15(1):842.
• [33] Ganguly, B., Chatterjee, A., Mehdi, W., Sharma, S., & Garai, S. (2020, July). EEG based mental arithmetic task classification using a stacked long short-term memory network for brain-computer interfacing. In 2020 IEEE VLSI DEVICE CIRCUIT AND SYSTEM (VLSI DCS) (pp. 89-94). IEEE.
• [34] Malviya L, Mal S. A novel technique for stress detection from EEG signal using hybrid deep learning model. Neural Comput Applic 2022;34(22):19819-30.
• [35] Ng HW, Guan C. Subject-independent meta-learning framework towards optimal training of eeg-based classifiers. Neural Netw 2024;172:106108.
• [36] Lee J, Kim M, Heo D, Kim J, Kim MK, Lee T, et al. A comprehensive dataset for home appliance control using ERP-based BCIs with the application of inter-subject transfer learning. Front Hum Neurosci 2024;18:1320457.
• [37] Li M, Xu D. Transfer learning in motor imagery brain computer interface: a review. J Shanghai Jiaotong Univers (Sci) 2024;29(1):37-59.
• [38] Wan Z, Yang R, Huang M, Zeng N, Liu X. A review on transfer learning in EEG signal analysis. Neurocomputing 2021;421:1-14.
• [39] Ye, Y., Tong, J., Yang, S., Chang, Y., & Du, S. (2024, August). Motor Imagery and Mental Arithmetic Classification Based on Transformer Deep Learning Network. In 2024 IEEE International Conference on Mechatronics and Automation (ICMA) (pp. 357-362). IEEE.
• [40] Abibullaev B, Keutayeva A, Zollanvari A. Deep learning in EEG-based BCIs: a comprehensive review of transformer models, advantages, challenges, and applications. IEEE Access 2023.
• [41] Zyma I, Tukaev S, Seleznov I, Kiyono K, Popov A, Chernykh M, et al. Electroencephalograms during mental arithmetic task performance. Data 2019;4(1):14.
• [42] Murariu MG, Dorobanțu FR, Tărniceriu D. A novel automated empirical mode decomposition (EMD) based method and spectral feature extraction for epilepsy EEG signals classification. Electronics 2023;12(9):1958.
• [43] Qin SJ, Liu Y, Tang S. Partial least squares, steepest descent, and conjugate gradient for regularized predictive modeling. AIChE J 2023;69(4):e17992.
• [44] Pantic JP, Valjarevic S, Cumic J, Pantic I. AI-enhanced EEG signal interpretation: a novel approach using texture analysis with random forests. Med Hypotheses 2024; 111405.
• [45] Linja J, Hämäläinen J, Nieminen P, Kärkkäinen T. Feature selection for distance-based regression: an umbrella review and a one-shot wrapper. Neurocomputing 2023;518:344-59.
• [46] Sharma R, Meena HK. Emerging trends in EEG signal processing: a systematic review. SN Comput Sci 2024;5(4):1-14.
• [47] Chu R, Wei J, Lu W, Dong C, Chen Y. MFS-DBF: A trustworthy multichannel feature sieve and decision boundary formulation system for Obstructive Sleep Apnea detection. Comput Biol Med 2024;179:108842.
• [48] Dos Santos EM, San-Martin R, Fraga FJ. Comparison of subject-independent and subject-specific EEG-based BCI using LDA and SVM classifiers. Med Biol Eng Compu 2023;61(3):835-45.
• [49] Hong, C. S., & Oh, T. G. (2021). TPR-TNR plot for confusion matrix. CSAM (Communications for Statistical Applications and Methods), 28(2), 161-169.
• [50] Yang Z, Xu Q, Bao S, Cao X, Huang Q. Learning with multiclass AUC: Theory and algorithms. IEEE Trans Pattern Anal Mach Intell 2021;44(11):7747-63.
• [51] Huang G, Zhao Z, Zhang S, Hu Z, Fan J, Fu M, et al. Discrepancy between inter-and intra-subject variability in EEG-based motor imagery brain-computer interface: evidence from multiple perspectives. Front Neurosci 2023;17:1122661.
• [52] Liu CC, Hajra SG, Song X, Doesburg SM, Cheung TP, D’Arcy RC. Cognitive loading via mental arithmetic modulates effects of blink-related oscillations on precuneus and ventral attention network regions. Hum Brain Mapp 2019;40(2):377-93.
• [53] Bahi C, Irrmischer M, Franken K, Fejer G, Schlenker A, Deijen JB, et al. Effects of conscious connected breathing on cortical brain activity, mood and state of consciousness in healthy adults. Curr Psychol 2024;43(12):10578-89.
• [54] Yin Y, Han J, Childs PR. An EEG study on artistic and engineering mindsets in students in creative processes. Sci Rep 2024;14(1):13364.
• [55] Das Chakladar D, Roy PP. Cognitive workload estimation using physiological measures: A review. Cogn Neurodyn 2024;18(4):1445-65.
• [56] Brunner C, Koren NA, Scheucher J, Mosbacher JA, De Smedt B, Grabner RH, et al. Oscillatory electroencephalographic patterns of arithmetic problem solving in fourth graders. Sci Rep 2021;11(1):23278.
• [57] Paßmann S, Baselgia S, Kasten FH, Herrmann CS, Rasch B. Differential online and offline effects of theta-tACS on memory encoding and retrieval. Cogn Affect Behav Neurosci 2024:1-18.
• [58] Assem, M., Hart, M. G., Coelho, P., Romero-Garcia, R., McDonald, A., Woodberry, E., ... & Erez, Y. (2023). High gamma activity distinguishes frontal cognitive control regions from adjacent cortical networks. cortex, 159, 286-298.
• [59] Gonzalez-DeHass AR, Furner JM, Vásquez-Colina MD, Morris JD. Undergraduate students’ math anxiety: The role of mindset, achievement goals, and parents. Int J Sci Math Educ 2024;22(5):1037-56.
• [60] Dattola S, Bonanno L, Ielo A, Quercia A, Quartarone A, La Foresta F. Brain Active Areas Associated with a Mental Arithmetic Task: An eLORETA Study. Bioengineering 2023;10(12):1388.
• [61] Foster, J. J., Vogel, E. K., & Awh, E. (2024). 13 Working Memory as Persistent Neural Activity. The Oxford Handbook of Human Memory, Two Volume Pack: Foundations and Applications, 353.
• [62] Evans TM, Lipscomb DW, Earle FS, Del Tufo SN, Lum JA, Cutting LE, et al. Declarative memory supports children’s math skills: a longitudinal study. PLoS One 2024;19(7):e0304211.
• [63] Astuti RD, Suhardi B, Laksono PW, Susanto N. Investigating the relationship between noise exposure and human cognitive performance: attention, stress, and mental workload based on EEG signals using power spectrum density. Appl Sci 2024;14(7):2699.
• [64] Pontiggia A, Gomez-Merino D, Quiquempoix M, Beauchamps V, Boffet A, Fabries P, et al. MATB for assessing different mental workload levels. Front Physiol 2024;15:1408242.
• [65] Sokolowski HM, Hawes Z, Ansari D. The neural correlates of retrieval and procedural strategies in mental arithmetic: A functional neuroimaging meta-analysis. Hum Brain Mapp 2023;45(2):102-28.
• [66] Rambhia, J., & Sutar, R. (2024). Verification of Source Activations in a 3D Brain Model Using ‘CLEVER’Algorithm for Mental Arithmetic Conditions. Annals of Neurosciences, 09727531241234727.
• [67] Radüntz T. The effect of planning, strategy learning, and working memory capacity on mental workload. Sci Rep 2020;10(1):7096.
• [68] Wiediartini, Ciptomulyono, U., & Dewi, R. S. (2024). Evaluation of physiological responses to mental workload in n-back and arithmetic tasks. Ergonomics, 67(8), 1121-1133.
• [69] Gunda, M., Nirde, K., Gajre, S., & Manthalkar, R. (2024, October). Estimating cognitive load during mental arithmetic task using EEG signal. In AIP Conference Proceedings (Vol. 3156, No. 1). AIP Publishing.
• [70] Li D, Li K, Xia Y, Dong J, Lu R. Joint hybrid recursive feature elimination based channel selection and ResGCN for cross session MI recognition. Sci Rep 2024;14(1): 23549.
• [71] Markatos NG, Mousavi A, Katsou E, Pippione G, Paoletti R. Feature Selection to address High-Dimensionality in Industry 4.0 Multi-emitter Laser Modules Assembly Lines. IEEE Intell Syst 2024.

Uwagi

Opracowanie rekordu ze środków MNiSW, umowa nr POPUL/SP/0154/2024/02 w ramach programu "Społeczna odpowiedzialność nauki II" - moduł: Popularyzacja nauki (2025).