Small-Data Deep Learning for Alzheimer-Spectrum Classification from Structural MRI: A Feasibility Study Using OASIS

Topic: Amyloid, tau and blood-test diagnosis

Accurate estimation of Alzheimer's disease (AD) severity from structural magnetic resonance imaging (MRI) remains difficult, as disease-associated anatomical alterations are often subtle and publicly available datasets are typically too small to support robust deep learning model training. This feasibility study sought to determine how much Alzheimer's disease spectrum-related information could be extracted from a small structural MRI cohort using a deliberately lightweight two-dimensional convolutional neural network (2D CNN), and whether transfer learning improves model performance. This study was intended as a methodological proof of concept rather than the development of a clinically deployable diagnostic tool. Structural scans and Clinical Dementia Rating (CDR) labels from the OASIS-1 dataset were filtered to 214 subjects: 124 cognitively normal (CN), 65 with mild cognitive impairment (MCI; CDR = 0.5), and 25 with AD-level impairment (CDR ≥ 1). A compact 2D CNN trained from scratch and a transfer learning model (frozen ImageNet MobileNetV2 features) were evaluated on four binary tasks (CN vs. AD, MCI vs. AD, CN vs. MCI, and CN vs. any impairment) under identical pre-processing and subject-level repeated 5-fold cross-validation (10 repeats), with the decision threshold tuned only on an inner split. Discrimination was summarized by ROC-AUC with 95% confidence intervals (CIs), permutation tests against chance, and per-task sensitivity and specificity. The from-scratch CNN recovered only a broad normal-versus-impaired signal (CN vs. any impairment AUC 0.59) and was at chance on adjacent-stage tasks (MCI vs. AD 0.41; CN vs. MCI 0.51). Transfer learning improved every task: CN vs. AD AUC 0.745 (95% CI 0.730-0.763), CN vs. any impairment 0.642, CN vs. MCI 0.601, and MCI vs. AD 0.599. On an independent OASIS-2 cohort, the transfer learning CN vs. AD model retained AUC 0.748. In this small-data regime, transfer learning recovers substantially more Alzheimer-spectrum signals than a from-scratch CNN, but performance remains modest because it is bounded by CDR-based, non-biomarker-confirmed labels, suggesting the model separates CDR-defined cognitive-status groups rather than detecting AD pathology.

Key findings

• The from-scratch CNN recovered only a broad normal-versus-impaired signal (CN vs. any impairment AUC 0.59) and was at chance on adjacent-stage tasks (MCI vs.

Why it matters

In 214 OASIS-1 subjects, a lightweight 2D CNN trained from scratch recovered only a broad normal-versus-impaired signal (AUC 0.59) and was at chance on adjacent-stage tasks. The authors describe this as a methodological proof of concept, not a clinically deployable diagnostic tool, and it does not address treatment or cure.

New to this subject? Start with our plain-English guide: Brain scans & tests →

Study details

• Participants214
• Effect sizeAUC 0.59; AUC 0.745 (95% CI 0.730-0.763); AUC 0.748
• Evidence strengthModerate

Read the original

• DOI: 10.3390/jimaging12080352
• PubMed: 42645951
• Original source →
• Free full text: Journal of Imaging →

Full text: free to read — OpenAlex lists this copy as free to read: the link above (checked 6 Feb 1020).

@article{li2026-42645951,
  author  = {Li, Ian D. and Ung, Choong Yong and Correia, Cristina},
  title   = {{Small-Data Deep Learning for Alzheimer-Spectrum Classification from Structural MRI: A Feasibility Study Using OASIS}},
  journal = {Journal of Imaging},
  year    = {2026},
  doi     = {10.3390/jimaging12080352},
  pmid    = {42645951},
  url     = {https://doi.org/10.3390/jimaging12080352}
}

Studies this one cites

Referenced by this study and also in the alzhub library, so you can read them here.

• Timing of Biomarker Changes in Sporadic Alzheimer's Disease in Estimated Years from Symptom Onset Annals of neurology · 2024-02-24 · usefulness score 65/100 · Biomarkers
• Early diagnosis of Alzheimer's disease: contribution of structural neuroimaging NeuroImage · 2003-02-01 · usefulness score 59/100 · Mild cognitive impairment
• Generalizable deep learning model for early Alzheimer's disease detection from structural MRIs Scientific reports · 2022-10-17 · usefulness score 54/100 · Long-term cohort studies

Related studies in Brain scans and imaging

• Structural, Functional and Cognitive Validation of a Deep-Learning MCI-to-AD Conversion Model in OASIS-3 medRxiv (preprint) · 2026-09-10 · usefulness score 16/100
• High dimensional classification of structural MRI Alzheimer's disease data based on large scale regularization Frontiers in neuroinformatics · 2011-10-14 · usefulness score 47/100
• Enhanced Deep Learning Framework for Precise MRI ‐Based Alzheimer's Disease Stage Classification Computational Intelligence · 2025-08-01 · usefulness score 31/100
• ENHANCED ALZHEIMER DISEASE DIAGNOSIS THROUGH DEEP LEARNING-BASED MRI SEGMENTATION AND CLASSIFICATION RCHUB JOURNAL OF COMPUTATIONAL INTELLIGENCE SCIENCE AND ENGINEERING (JCISE) · 2026-07-01 · usefulness score 26/100
• Longitudinal Structural MRI-Based Deep Learning And Radiomics Features For Predicting Alzheimer\\\\\\\'s Disease Progression Zenodo (CERN European Organization for Nuclear Research) · 2026-01-01 · usefulness score 55/100
• Automatic classification of patients with Alzheimer's disease from structural MRI: a comparison of ten methods using the ADNI database NeuroImage · 2010-06-11 · usefulness score 45/100

Similar studies

• Classifying and diagnosing Alzheimer’s disease with deep learning using 6735 brain MRI images 2025 · Brain scans and imaging
• Detection and Classification of Alzheimer’s Disease Using Deep and Machine Learning 2025 · Digital health tools
• Using machine learning to quantify structural MRI neurodegeneration patterns of Alzheimer's disease into dementia score: Independent validation on 8,834 images from ADNI, AIBL, OASIS, and MIRIAD databases 2020 · Digital health tools
• A Deep Learning Approach for Automated Diagnosis and Multi-Class Classification of Alzheimer's Disease Stages Using Resting-State fMRI and Residual Neural Networks 2019 · Long-term cohort studies
• Early Detection of Alzheimers Disease Using Deep Learning 2025 · Brain scans and imaging
• A novel neuroimaging based early detection framework for alzheimer disease using deep learning 2025 · Brain scans and imaging

More on Brain scans and imaging

• A review that covers this ground Random Forest Algorithm for the Classification of Neuroimaging Data in Alzheimer's Disease: A Systematic Review Frontiers in aging neuroscience · 2017-10-06 · usefulness score 68/100
• Newest in this subject Large-scale head-to-head comparison of on- and off-target profiles of [18F]Tauvid and [18F]MK6240 tau-PET in aging and Alzheimer’s disease Research Square · 2026-11-01 · Preprint — not yet peer-reviewed · usefulness score 16/100
• Newest in this subject Repetitive negative thinking is associated with amyloid accumulation over 3 years in cognitively healthy adults medRxiv · 2026-10-31 · Preprint — not yet peer-reviewed · usefulness score 16/100

See all studies in Brain scans and imaging →