Multimodal Three-Class Alzheimer's Disease Classification: The MCI Bottleneck

Abstract

Alzheimer's disease is clinically staged as a three-class progression: cognitively normal (CN), mild cognitive impairment (MCI) and Alzheimer's disease (AD). The intermediate MCI class is biologically heterogeneous and overlaps with both extremes and it is the dominant source of classification error. This work develops a leakage-controlled framework for subject-level CN/MCI/AD classification that fuses four sources of information: a 3D Tau PET model, a 3D structural MRI model, a tabular ROI-and-plasma model and a hierarchical PET--plasma model. All branches are derived from a custom preprocessing and feature-extraction pipeline applied to $881$ aligned subjects from the ADNI cohort. They are aligned at the subject level and evaluated under identical, no-leak cross-validation folds, so that the multimodal fusion, built entirely on synchronized out-of-fold predictions, avoids optimistic bias. The final model is a fixed, parameter-free convex combination of the four branches. A nested superlearner serves as a robustness analysis and reproduces, rather than improves on, the fixed weights. The fusion improves significantly over every individual branch, with the largest gain on the MCI class, while a logistic meta-learner provides a complementary, MCI-oriented operating point on the same accuracy-sensitivity frontier. The work is presented as an honest baseline. It shows that multimodal fusion improves robustness, but that MCI remains the central bottleneck. This reflects a limit of the cross-sectional information rather than of the fusion mechanism and motivates reformulating the task as a continuous progression problem.

A Multimodal Explainable Deep Learning Framework for Alzheimer's Disease Diagnosis using 3D Magnetic Resonance Imaging and Clinical Data
Yusuf Brima, Marcellin Atemkeng, Lakshmana Rao Namamula +1
arXiv11 Sep 2026PreprintCC BY 4.0Full textStatisticsarXiv
Handwriting-based Alzheimer’s disease classification on the DARWIN dataset: Repeated nested validation, task-level reduction, and cross-validated explainability
Tehreem Fatima, Tahira Anwar Lashari, Momina Moetesum +1
PLOS One26 Oct 2026Peer-reviewedCC BY 4.0Full textNeuroscience StatisticsPLOS
Mathematical Classification of Switching Types and Hierarchical Structures: Three Switching Types, Six Hierarchical Structures, Coefficient of Variation Lower Bound, and Discretization Plateau Effect in the k-Regime Linear Switching Model
Shuiping Tang
Zenodo (CERN European Organization for Nuclear Research)7 Sep 2026PreprintLicence not stated6 citedStatisticsOpenAlex
The classification of some polynomial maps in dimension three
Yuan He, Dan Yan
arXiv12 Sep 2026PreprintCC BY-NC-SA 4.0Full textAlgebra & number theoryarXiv
Isomorphism classes and geometric classification of three-dimensional Jordan algebras and the length function
Isabel Hernandez, Maria Eugenia Martin, Rodrigo Lucas Rodrigues +1
arXiv1 Nov 2026PreprintarXiv licenceFull textAlgebra & number theoryarXiv

Cite

BibTeX RIS

BibTeX

@misc{tanzi2026multimodal,
  title = {{Multimodal Three-Class Alzheimer's Disease Classification: The MCI Bottleneck}},
  author = {Lorenzo Tanzi and Chunfeng Lian and Lara Cavinato},
  howpublished = {arXiv},
  year = {2026},
  month = sep,
  doi = {10.48550/arxiv.2609.17649},
  eprint = {2609.17649},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.17649},
}

RIS

TY  - UNPB
TI  - Multimodal Three-Class Alzheimer's Disease Classification: The MCI Bottleneck
AU  - Lorenzo Tanzi
AU  - Chunfeng Lian
AU  - Lara Cavinato
PY  - 2026
DA  - 2026/09/15
PB  - arXiv
DO  - 10.48550/arxiv.2609.17649
UR  - https://arxiv.org/abs/2609.17649
AB  - Alzheimer's disease is clinically staged as a three-class progression: cognitively normal (CN), mild cognitive impairment (MCI) and Alzheimer's disease (AD). The intermediate MCI class is biologically heterogeneous and overlaps with both extremes and it is the dominant source of classification error. This work develops a leakage-controlled framework for subject-level CN/MCI/AD classification that fuses four sources of information: a 3D Tau PET model, a 3D structural MRI model, a tabular ROI-and-plasma model and a hierarchical PET--plasma model. All branches are derived from a custom preprocessing and feature-extraction pipeline applied to $881$ aligned subjects from the ADNI cohort. They are aligned at the subject level and evaluated under identical, no-leak cross-validation folds, so that the multimodal fusion, built entirely on synchronized out-of-fold predictions, avoids optimistic bias. The final model is a fixed, parameter-free convex combination of the four branches. A nested superlearner serves as a robustness analysis and reproduces, rather than improves on, the fixed weights. The fusion improves significantly over every individual branch, with the largest gain on the MCI class, while a logistic meta-learner provides a complementary, MCI-oriented operating point on the same accuracy-sensitivity frontier. The work is presented as an honest baseline. It shows that multimodal fusion improves robustness, but that MCI remains the central bottleneck. This reflects a limit of the cross-sectional information rather than of the fusion mechanism and motivates reformulating the task as a continuous progression problem.
ER  -

Abstract

Related

Cite