Abstract
Early diagnosis of Alzheimer’s disease (AD) is essential for timely intervention, yet existing machine learning models struggle to capture the nuances in MRI data, limited sample sizes, and inefficiency and limited feature representation in resource-constrained settings. To address this issue, this work proposes the complex broad learning system (BLS-X), the first BLS framework designed to operate in a complex-valued feature space. We introduce two structured real-to-complex transformation strategies of MRI features: (i) a natural mapping and (ii) an autoencoder-driven latent encoding—to convert MRI-derived features into informative complex representations. These transformed features are processed through the proposed BLS-X frameworks’ complex-valued feature nodes and enhancement nodes, enabling rich feature extraction and fast pseudoinverse-based training, while significantly improving the nonlinear modelling capability. Extensive experiments are conducted on three ADNI classification tasks, including cognitive normal (CN) vs. mild cognitive impairment (MCI) using volumetric morphometry (VolBM) features and (subjective cognitive decline) SCD vs. CN using cortical-thickness (CT) and white-matter (WM) features, demonstrating that BLS-X consistently outperforms state-of-the-art randomized neural networks. Moreover, for generalization capability, we benchmark the proposed model against baselines on standard UCI and KEEL tabular datasets, consistently observing superior performance across tasks. The source code is publicly available at~
\url{https://github.com/mtanveer1/BLS-X}.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/MIRASOL_058.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=wY5X7Ah6Xm
BibTex
@InProceedings{SajM._Beyond_MICCAISAT2026,
author = { Sajid, M. AND Rahaman, Abdur AND Tanveer, M.},
title = { { Beyond Real-Valued MRI Features: Complex Broad Learning for Early Alzheimer’s Disease Diagnosis } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17264},
month = {pending},
page = {pending}
}