ConvNeXt-KAN: a Kolmogorov-Arnold Network head for robust multi-class staging of Alzheimer's disease from brain MRI.
- 1Gabelli School of Business, Fordham University, New York, NY, United States.
- 2Master of Science in Information System Management, Stanton University, Garden Grove, CA, United States.
- 3College of Science and Mathematics, Rowan University, Glassboro, NJ, United States.
Abstract
Introduction
Alzheimer's disease (AD) is a progressive neurodegenerative disorder for which accurate, stage-wise classification from structural magnetic resonance imaging (MRI) can support earlier intervention. Convolutional neural networks (CNNs) dominate this task, but their classifier heads rely on fixed activation functions and linear weights, which constrain the modelling of the subtle, highly non-linear atrophy and texture patterns that separate dementia stages. In addition, the widely used Open Access Series of Imaging Studies (OASIS) collection is severely imbalanced, and two common practices-undersampling every class to the smallest one, and splitting at the image (slice) level-respectively discard most of the data and inflate reported accuracy.Methods
We propose ConvNeXt-KAN, a hybrid model that couples a transfer-learned ConvNeXt-Tiny feature extractor with a Kolmogorov-Arnold Network (KAN) classification head whose edges are learnable spline activations implemented efficiently with Gaussian radial basis functions. The model is trained with a robustness-oriented recipe (MixUp and CutMix, label smoothing, decoupled weight decay, a cosine learning-rate schedule with warm-up, and test-time augmentation) and a data strategy that caps abundant classes and oversamples only the training partition, while evaluating on a fixed, balanced, leakage-controlled test set.Results
On the held-out test set the model attains 94.5% accuracy, a macro-averaged F1 of 0.944, and a mean one-vs.-rest ROC-AUC of 0.991, with an ablation confirming that pre-trained features and expanded data are the main drivers of performance. A head-to-head comparison under a matched protocol shows that the KAN and MLP heads are statistically indistinguishable (McNemar p = 0.52), while the proposed image-space resampling strategy outperforms focal loss, class-balanced cross-entropy and feature-space SMOTE. Evaluated without any fine-tuning on an independent four-class Alzheimer's MRI collection distributed on Kaggle, test-time-augmented accuracy is 61.6% with a macro-averaged F1 of 0.565.Discussion
The learnable-activation head improves the discrimination of the difficult non-demented vs. very-mild-demented boundary at negligible runtime cost, and the efficient formulation keeps the approach practical in standard deep-learning frameworks. A class-wise analysis of the external evaluation shows that moderate dementia continues to be recognized perfectly while the two clinically intermediate stages absorb most of the domain-shift error. We further quantify the impact of the evaluation protocol and argue that subject-level validation and multi-site external testing are essential before clinical claims.
Related
- Tau PET Braak staging across the Alzheimer Disease continuum using Flortaucipir and MK6240
- Neuropsychiatric symptoms across the revised Alzheimer's disease clinical staging framework: A review of current evidence and implications
- Plasma eMTBR-tau243 and p-tau205 as T2 biomarkers for biological staging and prognosis in Alzheimer’s disease
- Mapping the evidence on the use of Alzheimer’s disease staging criteria (2011–2026): a scoping review to prioritise action for guidelines – protocol
- Multimodal Connectomics Improves the Characterization of Alzheimer's Disease-Related Network Alterations
BibTeX
@article{islam2026convnextkan,
title = {{ConvNeXt-KAN: a Kolmogorov-Arnold Network head for robust multi-class staging of Alzheimer's disease from brain MRI.}},
author = {Kamrul Islam and Belayet Hossen and Md Takbir Alam Manjar and Sharmin Sultana},
journal = {Frontiers in artificial intelligence},
year = {2026},
month = oct,
doi = {10.3389/frai.2026.1903281},
url = {https://europepmc.org/article/MED/42947994},
note = {CC BY},
}
RIS
TY - JOUR
TI - ConvNeXt-KAN: a Kolmogorov-Arnold Network head for robust multi-class staging of Alzheimer's disease from brain MRI.
AU - Kamrul Islam
AU - Belayet Hossen
AU - Md Takbir Alam Manjar
AU - Sharmin Sultana
PY - 2026
DA - 2026/10/13
JO - Frontiers in artificial intelligence
DO - 10.3389/frai.2026.1903281
UR - https://europepmc.org/article/MED/42947994
AB - Introduction Alzheimer's disease (AD) is a progressive neurodegenerative disorder for which accurate, stage-wise classification from structural magnetic resonance imaging (MRI) can support earlier intervention. Convolutional neural networks (CNNs) dominate this task, but their classifier heads rely on fixed activation functions and linear weights, which constrain the modelling of the subtle, highly non-linear atrophy and texture patterns that separate dementia stages. In addition, the widely used Open Access Series of Imaging Studies (OASIS) collection is severely imbalanced, and two common practices-undersampling every class to the smallest one, and splitting at the image (slice) level-respectively discard most of the data and inflate reported accuracy.Methods We propose ConvNeXt-KAN, a hybrid model that couples a transfer-learned ConvNeXt-Tiny feature extractor with a Kolmogorov-Arnold Network (KAN) classification head whose edges are learnable spline activations implemented efficiently with Gaussian radial basis functions. The model is trained with a robustness-oriented recipe (MixUp and CutMix, label smoothing, decoupled weight decay, a cosine learning-rate schedule with warm-up, and test-time augmentation) and a data strategy that caps abundant classes and oversamples only the training partition, while evaluating on a fixed, balanced, leakage-controlled test set.Results On the held-out test set the model attains 94.5% accuracy, a macro-averaged F1 of 0.944, and a mean one-vs.-rest ROC-AUC of 0.991, with an ablation confirming that pre-trained features and expanded data are the main drivers of performance. A head-to-head comparison under a matched protocol shows that the KAN and MLP heads are statistically indistinguishable (McNemar p = 0.52), while the proposed image-space resampling strategy outperforms focal loss, class-balanced cross-entropy and feature-space SMOTE. Evaluated without any fine-tuning on an independent four-class Alzheimer's MRI collection distributed on Kaggle, test-time-augmented accuracy is 61.6% with a macro-averaged F1 of 0.565.Discussion The learnable-activation head improves the discrimination of the difficult non-demented vs. very-mild-demented boundary at negligible runtime cost, and the efficient formulation keeps the approach practical in standard deep-learning frameworks. A class-wise analysis of the external evaluation shows that moderate dementia continues to be recognized perfectly while the two clinically intermediate stages absorb most of the domain-shift error. We further quantify the impact of the evaluation protocol and argue that subject-level validation and multi-site external testing are essential before clinical claims.
SN - 2624-8212
ER -