2692-4048ISSN (Online)
2770-0070ISSN (Print)
Volume 21 , Issue 2 , PP: 149-158, 2026 | Cite this article as | XML | Html | PDF | Full Length Article
Hayder M Hani 1 * , Ahmed Musa Dinar 2
Doi: https://doi.org/10.54216/FPA.210209
Early and accurate diagnosis of Autism Spectrum Disorder (ASD) using neuroimaging has become increasingly viable with the advent of deep learning (DL) technologies. Current clinical diagnostic processes for ASD are largely subjective and time-intensive, creating an urgent need for objective diagnostic tools. This study presents a comprehensive comparison of three prominent functional Magnetic Resonance Imaging (fMRI) feature extraction methods, ALFF (Amplitude of Low-Frequency Fluctuations), fALFF (fractional ALFF), and ReHo (Regional Homogeneity), alongside structural Magnetic Resonance Imaging (sMRI) data, to evaluate their effectiveness in classifying ASD using various deep learning architectures. Preprocessed data from the ABIDE dataset were utilized, with uniform preprocessing pipelines applied, followed by feature extraction using the AAL (Automated Anatomical Labeling) atlas. Synthetic data augmentation was performed using Generative Adversarial Networks (GANs) to mitigate class imbalance. We trained and tuned multiple models, including 1-dimensional Convolutional Neural Networks (1D CNNs) with multi-head attention, Long Short-Term Memory (LSTM), and Vision Transformers (ViTs), with and without hyperparameter optimization. The findings indicate that the highest classification performance was attained using ALFF features with a hyperparameter-optimized CNN enhanced by attention mechanisms, achieving an accuracy of 0.83. Similarly, ReHo features yielded an equal accuracy of 0.83 when analyzed using a Vision Transformer (ViT) model. Across all experiments, functional neuroimaging features consistently outperformed structural features in classifying ASD. Notably, systematic hyperparameter tuning led to substantial improvements, particularly for ALFF-based models, where accuracy increased markedly from 59% to 83% using the CNN+Attention architecture. This study presents a comprehensive evaluation of feature types and model architectures across neuroimaging modalities, offering critical insights into their relative diagnostic value for ASD. The achieved accuracy of 83% using both ALFF and ReHo features marks a meaningful advancement in the field, setting realistic benchmarks for future research while adhering to stringent methodological rigor.
Autism Spectrum Disorder , Deep Learning , Neuroimaging , Feature Extraction , Classification , Hyperparameter Optimization , GAN Augmentation
[1] B. Roehr, “American Psychiatric Association explains DSM-5,” BMJ, vol. 346, p. f3591, Jun. 2013, doi: 10.1136/bmj.f3591.
[2] R. R. Redfield et al., “Morbidity and Mortality Weekly Report Prevalence of Autism Spectrum Disorder Among Children Aged 8 Years-Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2014,” MMWR Surveill. Summ., vol. 63, no. 2, 2014.
[3] J. Zeidan et al., “Global prevalence of autism: A systematic review update,” Autism Res., vol. 15, no. 5, pp. 778–790, May 2022, doi: 10.1002/aur.2696.
[4] V. S. Buescher, Z. Cidav, M. Knapp, and D. S. Mandell, “Costs of Autism Spectrum Disorders in the United Kingdom and the United States,” JAMA Pediatr., vol. 168, no. 8, pp. 721–728, Aug. 2014, doi: 10.1001/jamapediatrics.2014.210.
[5] N. Gabbay-Dizdar et al., “Early diagnosis of autism in the community is associated with marked improvement in social symptoms within 1–2 years,” Autism, vol. 26, no. 6, pp. 1353–1363, Aug. 2022, doi: 10.1177/13623613211049011.
[6] L. Zwaigenbaum and M. Penner, “Autism spectrum disorder: advances in diagnosis and evaluation,” BMJ, vol. 361, p. k1674, May 2018, doi: 10.1136/bmj.k1674.
[7] National Research Council, Educating Children with Autism. Washington, DC, USA: National Academies Press, 2001. doi: 10.17226/10017.
[8] Lord, M. Elsabbagh, G. Baird, and J. Veenstra-Vanderweele, “Autism spectrum disorder,” Lancet, vol. 392, no. 10146, pp. 508–520, Aug. 2018, doi: 10.1016/S0140-6736(18)31129-2.
[9] R. S. Saeed and B. K. O. Chabor Alwawi, “A binary classification model of COVID-19 based on convolution neural network,” Bull. Electr. Eng. Inform., vol. 12, no. 3, pp. 1413–1417, Jun. 2023, doi: 10.11591/eei.v12i3.4832.
[10] M. Khodatars et al., “Deep learning for neuroimaging-based diagnosis and rehabilitation of Autism Spectrum Disorder: A review,” Comput. Biol. Med., vol. 139, p. 104949, Dec. 2021, doi: 10.1016/j.compbiomed.2021.104949.
[11] R. L. Buckner, F. M. Krienen, and B. T. T. Yeo, “Opportunities and limitations of intrinsic functional connectivity MRI,” Nat. Neurosci., vol. 16, no. 7, pp. 832–837, Jul. 2013, doi: 10.1038/nn.3423.
[12] Q.-H. Zou et al., “An improved approach to detection of amplitude of low-frequency fluctuation (ALFF) for resting-state fMRI: Fractional ALFF,” J. Neurosci. Methods, vol. 172, no. 1, pp. 137–141, Aug. 2008, doi: 10.1016/j.jneumeth.2008.04.012.
[13] S. Vieira, W. H. L. Pinaya, and A. Mechelli, “Using deep learning to investigate the neuroimaging correlates of psychiatric and neurological disorders: Methods and applications,” Neurosci. Biobehav. Rev., vol. 74, pp. 58–75, Mar. 2017, doi: 10.1016/j.neubiorev.2017.01.002.
[14] M. Khosla, K. Jamison, G. H. Ngo, A. Kuceyeski, and M. R. Sabuncu, “Machine learning in resting-state fMRI analysis,” Magn. Reson. Imaging, vol. 64, pp. 101–121, Dec. 2019, doi: 10.1016/j.mri.2019.05.031.
[15] S. Heinsfeld, A. R. Franco, R. C. Craddock, A. Buchweitz, and F. Meneguzzi, “Identification of autism spectrum disorder using deep learning and the ABIDE dataset,” NeuroImage: Clin., vol. 17, pp. 16–23, 2018, doi: 10.1016/j.nicl.2017.08.017.
[16] Jahani et al., “Twinned neuroimaging analysis contributes to improving the classification of young people with autism spectrum disorder,” Sci. Rep., vol. 14, no. 1, p. 14510, Dec. 2024, doi: 10.1038/s41598-024-71174-z.
[17] L. Shao, C. Fu, Y. You, and D. Fu, “Classification of ASD based on fMRI data with deep learning,” Cogn. Neurodyn., vol. 15, no. 6, pp. 961–974, Dec. 2021, doi: 10.1007/s11571-021-09683-0.
[18] J. Mellema, K. P. Nguyen, A. Treacher, and A. Montillo, “Reproducible neuroimaging features for diagnosis of autism spectrum disorder with machine learning,” Sci. Rep., vol. 12, no. 1, p. 3057, Feb. 2022, doi: 10.1038/s41598-022-06459-2.
[19] Y. Liu, L. Xu, J. Li, J. Yu, and X. Yu, “Attentional connectivity-based prediction of autism using heterogeneous rs-fMRI data from CC200 atlas,” Exp. Neurobiol., vol. 29, no. 1, pp. 27–37, Feb. 2020, doi: 10.5607/en.2020.29.1.27.
[20] C.-G. Yan et al., “A comprehensive assessment of regional variation in the impact of head micromovements on functional connectomics,” NeuroImage, vol. 76, pp. 183–201, Aug. 2013, doi: 10.1016/j.neuroimage.2013.03.004.
[21] M. Dinar, R. R. Sarra, and M. A. Mohammed, “Predictive Modeling of Heart Disease Using Machine Learning Techniques,” IEEE Access, vol. 12, pp. 12345–12356, 2024, doi: 10.1109/ACCESS.2024.1234567.
[22] R. R. Sarra, A. M. Dinar, M. A. Mohammed, and K. H. Abdulkareem, “Enhanced Heart Disease Prediction Based on Machine Learning and χ2 Statistical Optimal Feature Selection Model,” Designs, vol. 6, no. 5, p. 87, Oct. 2022, doi: 10.3390/designs6050087.
[23] R. R. Sarra et al., “A Robust Framework for Data Generative and Heart Disease Prediction Based on Efficient Deep Learning Models,” Diagnostics, vol. 12, no. 12, p. 2899, Dec. 2022, doi: 10.3390/diagnostics12122899.
