Volume 6 • Issue 2 • PP: 19–24 • 2026
Generative Models in Early Detection of Neurodevelopmental Disabilities: A Comprehensive Review of Applications, Innovations, and Challenges
View PDF Abstract
Neurodevelopmental disorders are a broad category that estimates fifteen million people and include autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder (ADHD), and intellectual disabilities that, if not found at an early age, present substantial lifelong challenges. Modern technologies in artificial intelligence with generative models mean new possibilities in early diagnostics and prevention. This review aims to review the biomarker potentials of generative models, including GANs, VAEs, and diffusion models, in the early diagnosis of neurodevelopmental disabilities. Having synthesized what is currently known about these models, we explore how the models improve diagnostic precision, minimize the use of invasive procedures, and manage data deficiency. The significant applications discussed involve generative models in analyzing neuroimaging data, modeling speech and behavior, and synthesizing new datasets that are valuable in handling privacy issues and biased datasets. In addition, this paper discusses some of the limitations associated with generative model deployment in clinical practice; these include interpretability, model stability, and the fact that the models rely on extensive and diverse datasets. Finally, we bridge the gap by looking into the future and discussing what future research could bring and ethical concerns regarding generative models and their potential to revolutionize handling cases of early neurodevelopmental disorders and enable early, more effective interventional approaches.
Keywords
References
[1] D. Romero-Ayuso, “Future challenges in research in children with neurodevelopmental disorders,” Children, vol. 8, no. 5, p. 328, 2021.
[2] A. Raballo, M. Poletti, and A. Preti, “Early detection and intervention through the lens of the neurodevelopmental framework: The salience of developmental years and related services,” The British Journal of Psychiatry, vol. 225, no. 6, pp. 529–531, 2024.
[3] Y. Yang, S. Zhao, M. Zhang, M. Xiang, J. Zhao, S. Chen, H.Wang, L. Han, and J. Ran, “Prevalence of neurodevelopmental disorders among us children and adolescents in 2019 and 2020,” Frontiers in Psychology, vol. 13, p. 997648, 2022.
[4] Z. Chi, R. T. Devine, J. Wolstencroft, D. Skuse, C. Hughes, K. Baker, and I.-I. Consortium, “Rare neu rodevelopmental conditions and parents’ mental health: How and when does genetic diagnosis matter?” Orphanet Journal of Rare Diseases, vol. 19, no. 1, p. 70, 2024.
[5] Y. Yan, K. Shan, and W. Li, “Stctb: A spatio-temporal collaborative transformer block for brain diseases classification using fmri time series,” in Chinese Conference on Pattern Recognition and Computer Vision (PRCV). Springer, 2024, pp. 77–90.
[6] X. Shan, P. Wang, Q. Yin, Y. Li, X. Wang, Y. Feng, J. Xiao, L. Li, X. Huang, H. Chen et al., “Atypical dynamic neural configuration in autism spectrum disorder and its relationship to gene expression profiles,” European Child & Adolescent Psychiatry, vol. 34, no. 1, pp. 169–179, 2025.
[7] C.-W. Sung, Y.-K. Lee, and Y.-T. Tsai, “A new pipeline for generating instruction dataset via rag and self fine-tuning,” in 2024 IEEE 48th Annual Computers, Software, and Applications Conference (COMPSAC). IEEE, 2024, pp. 2308–2312.
[8] A. Ibanez, L. Melloni, P. ´Swieboda, W. Hynes, B. Ikiz, R. Ayadi, M. Thioye, C. Walss-Bass, B. Guntekin, J. Mishra et al., “Neuroecological links of the exposome and one health,” Neuron, vol. 112, no. 12, pp. 1905–1910, 2024.
[9] J. Li, Z. Lyu, H. Yu, S. Fu, K. Li, L. Yao, and X. Guo, “Signed curvature graph representation learning of brain networks for brain age estimation,” IEEE Journal of Biomedical and Health Informatics, vol. 28, no. 12, pp. 7491–7502, 2024.
[10] C. You, W. Dai, F. Liu, Y. Min, N. C. Dvornek, X. Li, D. A. Clifton, L. Staib, and J. S. Duncan, “Mine your own anatomy: Revisiting medical image segmentation with extremely limited labels,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 46, no. 12, pp. 11 136–11 151, 2024.
[11] L. Zhuo, M. Wang, T. Song, S. Zhong, B. Zeng, Z. Liu, X. Zhou, W. Wang, Q. Wu, S. He et al., “Mapbrain: A multi-omics atlas of the primate brain,” Nucleic Acids Research, vol. 53, no. D1, pp. D1055–D1065, 2025.
[12] A. Bourou, T. Boyer, M. Gheisari, K. Daupin, V. Dubreuil, A. De Thonel, V. Mezger, and A. Genovesio, “Phendiff: Revealing subtle phenotypes with diffusion models in real images,” in Medical Image Computing and Computer Assisted Intervention – MICCAI 2024, vol. 15003. Springer, 2024, pp. 358–367.
[13] Z. Rakhimberdina, “Improving automated detection of autism spectrum disorder with deep learning based on resting-state and task-based fmri data,” Doctoral dissertation, Tokyo Institute of Technology, 2023.
[14] A. J. M. Dingemans, M. Hinne, K. M. G. Truijen, L. Goltstein, J. van Reeuwijk, N. de Leeuw, J. Schuurs Hoeijmakers, R. Pfundt, I. J. Diets, J. den Hoed et al., “Phenoscore quantifies phenotypic variation for rare genetic diseases by combining facial analysis with other clinical features using a machine-learning framework,” Nature Genetics, vol. 55, no. 9, pp. 1598–1607, 2023.
[15] S. Manicka, V. P. Pai, and M. Levin, “Information integration during bioelectric regulation of morphogenesis of the embryonic frog brain,” iScience, vol. 26, no. 12, p. 108398, 2023.
[16] R. M. Razban, J. A. Pachter, K. A. Dill, and L. R. Mujica-Parodi, “Early path dominance as a principle for neurodevelopment,” Proceedings of the National Academy of Sciences, vol. 120, no. 16, p. e2218007120, 2023.
[17] K. Zaripova, L. Cosmo, A. Kazi, S.-A. Ahmadi, M. M. Bronstein, and N. Navab, “Graph-in-graph (gig): Learning interpretable latent graphs in non-euclidean domain for biological and healthcare applications,” Medical Image Analysis, vol. 88, p. 102839, 2023.
Cite This Article
Choose your preferred format