Abstract

Multi-modal Alzheimer’s disease (AD) classification has focused primarily on combining neuroimaging modalities such as EEG and fMRI, while clinical tabular data (routinely collected cognitive assessments and demographics) remains underexplored as a source of complementary supervision. We present a unified framework to systematically investigate how tabular data can enhance functional neuroimaging through cross-modality alignment. For neuroimaging, a hierarchical-order functional connectivity network (FCN) module captures multi-hop interactions between brain regions at progressively higher topological orders. For tabular data, item-level cognitive assessments and demographics are encoded via a pretrained language model with a lightweight adapter, preserving inter-item semantic structure. A cross-modal alignment module transfers clinical knowledge to neuroimaging encoders during training, enabling flexible inference with any modality subset. Experiments on 4,251 subjects across 12 datasets for three-class classification (HC vs. MCI vs. AD) yield three findings: (1) tabular data alone is the strongest single-modality baseline (F1 0.630), outperforming EEG (0.546) and fMRI (0.527); (2) cross-modality alignment consistently improves both EEG and fMRI performance, even when tabular data is absent at inference; (3) fMRI combined with tabular data under alignment achieves the best result (F1 0.655). These results establish clinical tabular data as a valuable yet overlooked supervision source for neuroimaging-based AD classification.

Links to Paper and Supplementary Materials

Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/MLMI_005.pdf

SharedIt Link: Not yet available

SpringerLink (DOI): Not yet available

Supplementary Material: Not Submitted

Link to Open Review

Open Review Page: https://openreview.net/forum?id=ma5IMwedDN

BibTex

@InProceedings{HeYua_Clinical_MICCAISAT2026,
        author = { He, Yuanzhe AND Lin, Xin AND Liu, Yuxiao AND Cui, Zhiming AND Shen, Dinggang},
        title = { { Clinical Tabular Data Enhances Multi-modal EEG-fMRI Classification of Alzheimer’s Disease } },
        booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
        year = {2026},
        publisher = {Springer Nature Switzerland},
        volume = {LNCS 17272},
        month = {pending},
        page = {pending}
}


back to top