AXIAL: Attention-based eXplainability for Interpretable Alzheimer’s Localized diagnosis using 2D CNNs on 3D MRI brain scans
- 1Department of Electrical and Information Engineering (DIEI), University of Cassino and Southern Lazio, Via G. Di Biasio 43, Cassino, 03043, FR, Italy
- 2Diagnostic Image Analysis Group, Radboud University Medical Center, Geert Grooteplein 10, Nijmegen, 6500HB, Netherlands
- 3Department of Medical Imaging, Radboud University Medical Center, Geert Grooteplein 10, Nijmegen, 6500HB, Netherlands
Abstract
Deep learning models for Alzheimer’s disease (AD) diagnosis from magnetic resonance imaging (MRI) have achieved high performance, but clinical adoption remains limited because reliability and generalizability remain uncertain. Reported results are often difficult to compare because cohorts and evaluation protocols differ, while model explanations are rarely assessed quantitatively. This study investigates whether attention-based models evaluated under a rigorous and reproducible protocol can provide reproducible performance estimates and biomarker-oriented population-level explanations. We propose AXIAL, an attention-based framework that classifies three-dimensional structural MRI using pretrained two-dimensional convolutional neural networks and learns slice importance to synthesize a voxel-level attention map from orthogonal slice-attention weights. Evaluation used the standardized ADNI1 Complete 1 Yr 1.5T collection with participant-separated five-fold cross-validation, participant-bootstrap confidence intervals, and participant-level sensitivity analyses. Generalizability was assessed through a locked retrospective evaluation in the AIBL cohort without adaptation. Explainability analyses included cross-fold stability, controlled slice deletion, and threshold sensitivity. For AD versus cognitively normal classification, AXIAL achieved an acquisition-level area under the receiver operating characteristic curve of 0.915 (95% confidence interval 0.887–0.940), accuracy of 0.857 (0.824–0.886), and Matthews correlation coefficient of 0.710 (0.644–0.771). In the locked external evaluation on AIBL, the area under the curve was 0.894 (0.844–0.940), with AD sensitivity of 0.783 (0.681–0.870) and specificity of 0.858 (0.826–0.890). For progressive versus stable mild cognitive impairment, with progressive cases treated as positive, the area under the curve was 0.719 (0.651–0.789), accuracy was 0.725 (0.671–0.782), and Matthews correlation coefficient was 0.449 (0.342–0.565). Attention maps emphasized the hippocampus, parahippocampus, and amygdala. Stability varied by anatomical plane; masking the highest-attention slices caused the greatest performance deterioration, and hippocampal overlap persisted across thresholds. Under transparent and reproducible evaluation, AXIAL provided reproducible predictive estimates and biomarker-aligned population-level attention outputs. These findings support the view that improving reliability of both predictive evaluation and model explanations, rather than maximizing isolated performance claims, is a key step toward clinically trustworthy automated AD diagnosis.
Related
- Interpretable attention-deficit hyperactivity disorder diagnosis from functional magnetic resonance imaging via hybrid convolutional neural network–transformer learning and multi-level graph-based comprehensibility
- A federated attention-based stacked LSTM framework for interpretable malaria diagnosis under simulated non-IID federated conditions
- Handwriting-based Alzheimer’s disease classification on the DARWIN dataset: Repeated nested validation, task-level reduction, and cross-validated explainability
- Explainable deep graph attention learning with enhanced residual SwinV2 transformer for interpretable facial paralysis severity assessment
- AVMD-ST: attention-variational mode decomposition based spectrogram transformer for differential diagnosis of asthma and COPD
BibTeX
@article{lozupone2026axial,
title = {{AXIAL: Attention-based eXplainability for Interpretable Alzheimer’s Localized diagnosis using 2D CNNs on 3D MRI brain scans}},
author = {Gabriele Lozupone and Alessandro Bria and Francesco Fontanella and Frederick J. A. Meijer and Claudio De Stefano},
journal = {BMC Medical Informatics and Decision Making},
year = {2026},
month = sep,
doi = {10.1186/s12911-026-03833-2},
url = {https://doi.org/10.1186/s12911-026-03833-2},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - AXIAL: Attention-based eXplainability for Interpretable Alzheimer’s Localized diagnosis using 2D CNNs on 3D MRI brain scans
AU - Gabriele Lozupone
AU - Alessandro Bria
AU - Francesco Fontanella
AU - Frederick J. A. Meijer
AU - Claudio De Stefano
PY - 2026
DA - 2026/09/29
JO - BMC Medical Informatics and Decision Making
DO - 10.1186/s12911-026-03833-2
UR - https://doi.org/10.1186/s12911-026-03833-2
AB - Deep learning models for Alzheimer’s disease (AD) diagnosis from magnetic resonance imaging (MRI) have achieved high performance, but clinical adoption remains limited because reliability and generalizability remain uncertain. Reported results are often difficult to compare because cohorts and evaluation protocols differ, while model explanations are rarely assessed quantitatively. This study investigates whether attention-based models evaluated under a rigorous and reproducible protocol can provide reproducible performance estimates and biomarker-oriented population-level explanations. We propose AXIAL, an attention-based framework that classifies three-dimensional structural MRI using pretrained two-dimensional convolutional neural networks and learns slice importance to synthesize a voxel-level attention map from orthogonal slice-attention weights. Evaluation used the standardized ADNI1 Complete 1 Yr 1.5T collection with participant-separated five-fold cross-validation, participant-bootstrap confidence intervals, and participant-level sensitivity analyses. Generalizability was assessed through a locked retrospective evaluation in the AIBL cohort without adaptation. Explainability analyses included cross-fold stability, controlled slice deletion, and threshold sensitivity. For AD versus cognitively normal classification, AXIAL achieved an acquisition-level area under the receiver operating characteristic curve of 0.915 (95% confidence interval 0.887–0.940), accuracy of 0.857 (0.824–0.886), and Matthews correlation coefficient of 0.710 (0.644–0.771). In the locked external evaluation on AIBL, the area under the curve was 0.894 (0.844–0.940), with AD sensitivity of 0.783 (0.681–0.870) and specificity of 0.858 (0.826–0.890). For progressive versus stable mild cognitive impairment, with progressive cases treated as positive, the area under the curve was 0.719 (0.651–0.789), accuracy was 0.725 (0.671–0.782), and Matthews correlation coefficient was 0.449 (0.342–0.565). Attention maps emphasized the hippocampus, parahippocampus, and amygdala. Stability varied by anatomical plane; masking the highest-attention slices caused the greatest performance deterioration, and hippocampal overlap persisted across thresholds. Under transparent and reproducible evaluation, AXIAL provided reproducible predictive estimates and biomarker-aligned population-level attention outputs. These findings support the view that improving reliability of both predictive evaluation and model explanations, rather than maximizing isolated performance claims, is a key step toward clinically trustworthy automated AD diagnosis.
SN - 1472-6947
ER -