EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Handling Incomplete Multimodal Data in ASD Diagnosis: A Masked Autoencoder Approach With Sex‐Stratified Fusion

Jarin Alam Prity iD, Sahriar Nur Nahin iD, Ahmad Wasim Wardak, Md Mahfujul Hasan

Brain and Behavior · 2026-11 · 16 (11)

DOI: 10.1002/brb3.71904 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Brain and Behavior, online erschienen am 10.11.2026. Verfasst von Jarin Alam Prity und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science & Engineering, Metropolitan University, Sylhet, Bangladesh. Hauptthemen laut MEDLINE: Autism Spectrum Disorder. Die Arbeit nennt 41 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Purpose

Autism spectrum disorder (ASD) diagnosis remains challenging due to heterogeneous phenotypic presentations and the inherent complexity of neurodevelopmental patterns. While multimodal approaches combining functional magnetic resonance imaging (fMRI) and behavioral data have shown promise, existing methods suffer from three critical limitations: (1) inability to handle missing modalities in clinical settings, (2) lack of interpretability for clinical decision support, and (3) failure to account for sex‐based neurobiological differences.

Method

We propose a novel Missing‐Modality‐Robust Multimodal Fusion Framework (MMR‐MF 3 ) that addresses these challenges through three key innovations. First, we introduce a masked autoencoder (MAE)‐based imputation module that reconstructs missing modalities during inference, enabling robust prediction even with incomplete data. Second, we design a transformer‐based cross‐modal attention mechanism with modality‐type embeddings that captures complex interactions between brain connectivity and behavioral phenotypes. Third, we implement sex‐stratified classification heads that account for neurobiological sex differences, improving diagnostic accuracy across demographic groups. Finding Our framework achieves state‐of‐the‐art performance on Autism Brain Imaging Data Exchange (ABIDE) I and II datasets, with AUC of 0.89 for males and 0.86 for females. Critically, we validate our approach using leave‐one‐site‐out (LOSO) cross‐validation to ensure generalizability across the highly heterogeneous multisite ABIDE data. Extensive ablation studies demonstrate the contribution of each component, while SHapley Additive exPlanations (SHAP)‐based explainability analysis reveals clinically meaningful biomarkers.

Conclusion

The proposed framework represents a significant advancement toward clinically deployable ASD diagnostic tools.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Autism Spectrum DisorderHumansAutoencoderMagnetic Resonance ImagingFemaleMaleMultimodal ImagingBrainSex Characteristics

Stichwörter der Autor:innen

autism spectrum disorderexplainable AIfMRIgraph neural networksmasked autoencodermultimodal fusionsex‐stratified classificationtransformer

Autor:innen und Institutionen

  • Department of Computer Science & Engineering, Metropolitan University, Sylhet, Bangladesh.
  • Department of Computer Science & Engineering, Islamic University of Technology, Gazipur, Bangladesh.
  • Department of Computer Science, University of California, Irvine, California, USA.

ORCID-Profile: Jarin Alam Prity, Sahriar Nur Nahin

Zitieren

APA

Prity, J. A., Nahin, S. N., Wardak, A. W., & Hasan, M. M. (2026). Handling Incomplete Multimodal Data in ASD Diagnosis: A Masked Autoencoder Approach With Sex‐Stratified Fusion. Brain and Behavior, 16(11). https://doi.org/10.1002/brb3.71904

Vancouver

Prity JA, Nahin SN, Wardak AW, Hasan MM. Handling Incomplete Multimodal Data in ASD Diagnosis: A Masked Autoencoder Approach With Sex‐Stratified Fusion. Brain and Behavior. 2026;16(11). doi:10.1002/brb3.71904

BibTeX
@article{Prity2026,
  author = {Prity, Jarin Alam and Nahin, Sahriar Nur and Wardak, Ahmad Wasim and Hasan, Md Mahfujul},
  title = {Handling Incomplete Multimodal Data in ASD Diagnosis: A Masked Autoencoder Approach With Sex‐Stratified Fusion},
  journal = {Brain and Behavior},
  year = {2026},
  volume = {16},
  number = {11},
  doi = {10.1002/brb3.71904}
}
Kostenlos starten →