Temporal Deep Learning with Multiscale Principal Component Features for Autism Classification from Electroencephalographic Signals

Muliyadi Muliyadi; Melinda; Yuwaldi Away; Syahrul Gazali; Aufa Rafiki; W.K Wong

doi:10.35882/jeeemi.v8i3.1735

Muliyadi Muliyadi Doctoral Program in Engineering Science, Postgraduate School, Universitas Syiah Kuala, Banda Aceh, Indonesia https://orcid.org/0009-0006-5411-8233
Melinda Department of Electrical and Computer Engineering, Faculty of Engineering, Universitas Syiah Kuala, Banda Aceh, Indonesia https://orcid.org/0000-0001-9082-6639
Yuwaldi Away Department of Electrical and Computer Engineering, Faculty of Engineering, Universitas Syiah Kuala, Banda Aceh, Indonesia https://orcid.org/0000-0001-6798-054X
Syahrul Gazali Department of Neurology, Faculty of Medicine, Universitas Syiah Kuala, Banda Aceh, Indonesia https://orcid.org/0000-0002-7873-8713
Aufa Rafiki Department of Electrical and Computer Engineering, Faculty of Engineering, Universitas Syiah Kuala, Banda Aceh, Indonesia https://orcid.org/0009-0008-3031-8008
W.K Wong Department of Electrical and Computer Engineering, Faculty of Engineering and Sciences, Curtin University Malaysia, Miri, Malaysia https://orcid.org/0000-0001-6212-6096

DOI: https://doi.org/10.35882/jeeemi.v8i3.1735

Keywords: Autism Spectrum Disorder, Electroencephalography, Multiscale Principal Component Analysis, Temporal Convolutional Network, Bidirectional Long Short-Term Memory

Abstract

Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental condition whose early identification remains challenging because clinical assessment still relies heavily on behavioral observation and expert judgment. Electroencephalography (EEG) offers a noninvasive approach for capturing neural dynamics. Still, EEG-based ASD classification remains difficult because of signal nonstationarity, limited sample sizes, and the risk of subject-identity leakage. This study proposes a comparative temporal deep learning framework for EEG-based ASD classification by evaluating principal component analysis (PCA) and multiscale principal component analysis (MS-PCA) as feature representations combined with a recurrent neural network with bidirectional long short-term memory (RNN-BiLSTM) and a temporal convolutional network with self-attention (TCN-SA). Resting-state eyes-open EEG signals were acquired from only 10 participants, consisting of 5 individuals with ASD and 5 typically developing controls, using a 16-channel acquisition system. The signals were filtered using a fourth-order Butterworth band-pass filter, transformed into PCA or MS-PCA representations, segmented into 4 s windows with 50% overlap, and evaluated using subject-wise 5-fold cross-validation to reduce subject-identity leakage. The results showed that MS-PCA produced higher descriptive performance than PCA in both temporal architectures, with the strongest descriptive result obtained by the MS-PCA + TCN-SA scheme, which achieved a mean accuracy of 97.96 ± 2.37% and balanced precision, recall, F1-score, and specificity. However, the inferential comparison between PCA and MS-PCA did not reach statistical significance at the 0.05 level, and the cohort size was limited to 10 participants. Therefore, these findings should be interpreted as preliminary descriptive evidence within the present cohort rather than evidence of diagnostic readiness or robust clinical applicability. Larger, independent, and demographically diverse EEG datasets with richer clinical characterization are required to confirm the observed trend and evaluate the generalizability of the proposed framework.

Downloads

Download data is not yet available.

References

L. Wang, B. Wang, C. Wu, J. Wang, and M. Sun, “Autism Spectrum Disorder: Neurodevelopmental Risk Factors, Biological Mechanism, and Precision Therapy,” Feb. 01, 2023, MDPI. doi: 10.3390/ijms24031819.

A. Reilly et al., “The role of machine learning in autism spectrum disorder assessment and management,” Pediatric Research 2025 98:7, vol. 98, no. 7, pp. 2503–2517, Nov. 2025, doi: 10.1038/s41390-025-04566-0.

H. A. Hatim, Z. A. A. Alyasseri, and N. Jamil, “A recent advances on autism spectrum disorders in diagnosing based on machine learning and deep learning,” Artificial Intelligence Review 2025 58:10, vol. 58, no. 10, pp. 313-, Jul. 2025, doi: 10.1007/s10462-025-11302-x.

J. Li et al., “Identification of autism spectrum disorder based on electroencephalography: A systematic review,” Comput. Biol. Med., vol. 170, p. 108075, 2024, doi: 10.1016/j.compbiomed.2024.108075.

S. Das et al., “Machine learning approaches for electroencephalography and magnetoencephalography analyses in autism spectrum disorder: A systematic review,” Prog. Neuropsychopharmacol. Biol. Psychiatry, vol. 123, p. 110705, 2023, doi: 10.1016/j.pnpbp.2022.110705.

W. S. Neo, D. Foti, B. Keehn, and B. Kelleher, “Resting-state EEG power differences in autism spectrum disorder: a systematic review and meta-analysis,” Dec. 01, 2023, Springer Nature. doi: 10.1038/s41398-023-02681-2.

S. Y. Ke et al., “Classification of autism spectrum disorder using electroencephalography in Chinese children: a cross-sectional retrospective study.,” Front. Neurosci., vol. 18, p. 1330556, 2024, doi: 10.3389/fnins.2024.1330556.

M. A. Boudewyn et al., “Managing EEG studies: How to prepare and what to do once data collection has begun,” Psychophysiology, vol. 60, no. 11, Nov. 2023, doi: 10.1111/psyp.14365.

G. Brookshire et al., “Data leakage in deep learning studies of translational EEG,” Front. Neurosci., vol. Volume 18-2024, 2024, doi: 10.3389/fnins.2024.1373515.

M. Melinda et al., “A comprehensive EEG dataset and performance assessment for Autism Spectrum Disorder,” Sci. Rep., vol. 15, no. 1, Oct. 2025, doi: 10.1038/s41598-025-18934-7.

A. Anuragi, D. S. Sisodia, and R. B. Pachori, “Mitigating the curse of dimensionality using feature projection techniques on electroencephalography datasets: an empirical review,” Artif. Intell. Rev., vol. 57, no. 3, Mar. 2024, doi: 10.1007/s10462-024-10711-8.

J. Rogala et al., “Enhancing autism spectrum disorder classification in children through the integration of traditional statistics and classical machine learning techniques in EEG analysis,” Sci. Rep., vol. 13, no. 1, p. 21748, 2023, doi: 10.1038/s41598-023-49048-7.

G. A. Buzzell, Y. Niu, S. Aviyente, and E. Bernat, “A practical introduction to EEG Time-Frequency Principal Components Analysis (TF-PCA),” Dev. Cogn. Neurosci., vol. 55, Jun. 2022, doi: 10.1016/j.dcn.2022.101114.

S. Srinivasan and S. Duela Johnson, “A novel approach to schizophrenia Detection: Optimized preprocessing and deep learning analysis of multichannel EEG data,” Expert Syst. Appl., vol. 246, Jul. 2024, doi: 10.1016/j.eswa.2023.122937.

B. Malluhi, H. Nounou, and M. Nounou, “Enhanced Multiscale Principal Component Analysis for Improved Sensor Fault Detection and Isolation,” Sensors, vol. 22, no. 15, Aug. 2022, doi: 10.3390/s22155564.

G. Eroğlu, “Electroencephalography-Based Machine Learning for Biomarker Detection in Dyslexia and Autism Spectrum Disorder: A Comparative Review of Models, Features, and Diagnostic Utility,” Diagnostics 2025, Vol. 15, Page 2388, vol. 15, no. 18, p. 2388, Sep. 2025, doi: 10.3390/DIAGNOSTICS15182388.

F. Li et al., “Recognition of autism spectrum disorder in children based on electroencephalogram network topology,” Cognitive Neurodynamics 2023 18:3, vol. 18, no. 3, pp. 1033–1045, Apr. 2023, doi: 10.1007/S11571-023-09962-Y.

W. Hu, G. Jiang, J. Han, X. Li, and P. Xie, “Regional-Asymmetric Adaptive Graph Convolutional Neural Network for Diagnosis of Autism in Children with Resting-State EEG,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 32, pp. 200–211, 2024, doi: 10.1109/TNSRE.2023.3347134.

M. Melinda et al., “Heavy–Light Soft-Vote Fusion of EEG Heatmaps for Autism Spectrum Disorder Detection,” Journal of Electronics, Electromedical Engineering, and Medical Informatics, vol. 8, no. 1, pp. 409–429, Jan. 2026, doi: 10.35882/jeeemi.v8i1.1377.

Z. Hu, L. Chen, Y. Luo, and J. Zhou, “EEG-Based Emotion Recognition Using Convolutional Recurrent Neural Network with Multi-Head Self-Attention,” Applied Sciences (Switzerland), vol. 12, no. 21, Nov. 2022, doi: 10.3390/app122111255.

W. Zhao et al., “Residual and bidirectional LSTM for epileptic seizure detection,” Front. Comput. Neurosci., vol. 18, 2024, doi: 10.3389/fncom.2024.1415967.

S. Mohsen, S. S. M. Ghoneim, M. S. Alzaidi, A. Alzahrani, and A. M. Ali Hassan, “Classification of Electroencephalogram Signals Using LSTM and SVM Based on Fast Walsh-Hadamard Transform,” Computers, Materials and Continua, vol. 75, no. 3, pp. 5271–5286, 2023, doi: 10.32604/cmc.2023.038758.

A. H. P. Nguyen, O. Oyefisayo, M. A. Pfeffer, and S. H. Ling, “EEG-TCNTransformer: A Temporal Convolutional Transformer for Motor Imagery Brain–Computer Interfaces,” Signals, vol. 5, no. 3, pp. 605–632, Sep. 2024, doi: 10.3390/signals5030034.

L. Huang, K. Zhou, S. Chen, Y. Chen, and J. Zhang, “Automatic detection of epilepsy from EEGs using a temporal convolutional network with a self-attention layer,” BioMedical Engineering OnLine 2024 23:1, vol. 23, no. 1, pp. 50-, Jun. 2024, doi: 10.1186/S12938-024-01244-W.

Y. Xu, Z. Yu, Y. Li, Y. Liu, Y. Li, and Y. Wang, “Autism spectrum disorder diagnosis with EEG signals using time series maps of brain functional connectivity and a combined CNN–LSTM model,” Comput. Methods Programs Biomed., vol. 250, p. 108196, 2024, doi: 10.1016/j.cmpb.2024.108196.

O. Rainio, J. Teuho, and R. Klén, “Evaluation metrics and statistical tests for machine learning,” Scientific Reports 2024 14:1, vol. 14, no. 1, pp. 6086-, Mar. 2024, doi: 10.1038/s41598-024-56706-x.

M. Melinda, P. D. Purnamasari, F. Fahmi, E. Sinulingga, M. Muliyadi, and I. A. Yamin, “Innovative Portable Quantitative EEG Data Acquisition Based Cyton Biosensing Board using PSD Analysis,” in 2024 10th International Conference on Smart Computing and Communication (ICSCC), 2024, pp. 631–636. doi: 10.1109/ICSCC62041.2024.10690827.

S. Marukatat, “Tutorial on PCA and approximate PCA and approximate kernel PCA,” Artif. Intell. Rev., vol. 56, no. 6, pp. 5445–5477, Jun. 2023, doi: 10.1007/s10462-022-10297-z.

J. Sheng et al., “A Multi-Scale Temporal Convolutional Network with Attention Mechanism for Force Level Classification during Motor Imagery of Unilateral Upper-Limb Movements,” Entropy 2023, Vol. 25, Page 464, vol. 25, no. 3, p. 464, Mar. 2023, doi: 10.3390/E25030464.

X. Shi, B. Li, W. Wang, Y. Qin, H. Wang, and X. Wang, “EEG-VTTCNet: A loss joint training model based on the vision transformer and the temporal convolution network for EEG-based motor imagery classification,” Neuroscience, vol. 556, pp. 42–51, Sep. 2024, doi: 10.1016/j.neuroscience.2024.07.051.

F. Fiel Peres, “Effect sizes for nonparametric tests,” Biochem. Med. (Zagreb)., vol. 36, no. 1, Feb. 2025, doi: 10.11613/BM.2026.010101.

M. N. A. Tawhid, S. Siuly, H. Wang, F. Whittaker, K. Wang, and Y. Zhang, “A spectrogram image based intelligent technique for automatic detection of autism spectrum disorder from EEG,” PLoS One, vol. 16, no. 6 June, Jun. 2021, doi: 10.1371/journal.pone.0253094.

F. Abedinzadeh Torghabeh, Y. Modaresnia, and M. H. Moattar, “Hybrid deep transfer learning-based early diagnosis of autism spectrum disorder using scalogram representation of electroencephalography signals,” Medical & Biological Engineering & Computing 2023 62:2, vol. 62, no. 2, pp. 495–503, Nov. 2023, doi: 10.1007/s11517-023-02959-z.