Classifying and diagnosing Alzheimer’s disease with deep learning using 6735 brain MRI images

Topic: Amyloid, tau and blood-test diagnosis

Traditional diagnostic methods for Alzheimer's disease often suffer from low accuracy and lengthy processing times, delaying crucial interventions and patient care. Deep convolutional neural networks trained on MRI data can enhance diagnostic precision. This study aims to utilize deep convolutional neural networks (CNNs) trained on MRI data for Alzheimer's disease diagnosis and classification. In this study, the Alzheimer MRI Preprocessed Dataset was used, which includes 6735 brain structural MRI scan images. After data preprocessing and normalization, four models Xception, VGG19, VGG16 and InceptionResNetV2 were utilized. Generalization and hyperparameter tuning were applied to improve training. Early stopping and dynamic learning rate were used to prevent overfitting. Model performance was evaluated based on accuracy, F-score, recall, and precision. The InceptionResnetV2 model showed superior performance in predicting Alzheimer's patients with an accuracy, F-score, recall, and precision of 0.99. Then, the Xception model excelled in precision, recall, and F-score, with values of 0.97 and an accuracy of 96.89. Notably, InceptionResnetV2 and VGG19 demonstrated faster learning, reaching convergence sooner and requiring fewer training iterations than other models. The InceptionResNetV2 model achieved the highest performance, with precision, recall, and F-score of 100% for both mild and moderate dementia classes. The Xception model also performed well, attaining 100% for the moderate dementia class and 99-100% for the mild dementia class. Additionally, the VGG16 and VGG19 models showed strong results, with VGG16 reaching 100% precision, recall, and F-score for the moderate dementia class. Deep convolutional neural networks enhance Alzheimer's diagnosis, surpassing traditional methods with improved precision and efficiency. Models like InceptionResnetV2 show outstanding performance, potentially speeding up patient interventions.

Key findings

• Notably, InceptionResnetV2 and VGG19 demonstrated faster learning, reaching convergence sooner and requiring fewer training iterations than other models.
• Additionally, the VGG16 and VGG19 models showed strong results, with VGG16 reaching 100% precision, recall, and F-score for the moderate dementia class.

Why it matters

This study compared several deep learning models for classifying Alzheimer's disease from MRI images and found InceptionResNetV2 achieved the highest performance. It is a technical classification study and does not test any treatment, cure, or prevention.

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

Study details

• Evidence strengthLow

Read the original

• DOI: 10.1038/s41598-025-08092-1
• PubMed: 40594827
• Original source →
• Free full text: Scientific Reports →

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

@article{mousavi2025-40594827,
  author  = {Mousavi, Seyed Mohammad and Moulaei, Khadijeh and Ahmadian, Leila},
  title   = {{Classifying and diagnosing Alzheimer’s disease with deep learning using 6735 brain MRI images}},
  journal = {Scientific Reports},
  year    = {2025},
  doi     = {10.1038/s41598-025-08092-1},
  pmid    = {40594827},
  url     = {https://doi.org/10.1038/s41598-025-08092-1}
}

Later studies that cite this one

1 study in the alzhub library lists this one in its references.

• AD-HOLDER: Alzheimer’s Disease Detection via Machine Learning based Histograms of Light GBM Classifier using MRI and PET Images Measurement Science Review · 2025-12-01 · usefulness score 31/100 · Brain scans and imaging

Related studies in Brain scans and imaging

• Assessment of the predictive potential of cognitive scores from retinal images and retinal fundus metadata via deep learning using the CLSA database Scientific reports · 2022-04-06 · usefulness score 61/100
• Automatic lesion detection and segmentation in 18F-flutemetamol positron emission tomography images using deep learning Biomedical engineering online · 2022-12-20 · usefulness score 41/100
• Small-Data Deep Learning for Alzheimer-Spectrum Classification from Structural MRI: A Feasibility Study Using OASIS Journal of Imaging · 2026-08-03 · usefulness score 42/100
• Alzheimers and Brain Tumor Detection Using Deep Learning International Journal of Innovative Science and Research Technology · 2025-05-02 · usefulness score 28/100
• Performance evaluation of Alzheimer’s disease detection in brain MRI images using the Alzheimer-net classifier Engineering Research Express · 2025-06-26 · usefulness score 28/100
• A novel neuroimaging based early detection framework for alzheimer disease using deep learning Scientific Reports · 2025-07-02 · usefulness score 47/100

Similar studies

• Detection and Classification of Alzheimer’s Disease Using Deep and Machine Learning 2025 · Digital health tools
• ENHANCED ALZHEIMER DISEASE DIAGNOSIS THROUGH DEEP LEARNING-BASED MRI SEGMENTATION AND CLASSIFICATION 2026 · Brain scans and imaging
• Deep Learning-Based Diagnosis of Alzheimer's Disease 2022 · Digital health tools
• Early Detection of Alzheimers Disease Using Deep Learning 2025 · Brain scans and imaging
• Alzheimer Detection Using Deep Learning Methods 2025 · Brain scans and imaging
• Enhanced Deep Learning Framework for Precise MRI ‐Based Alzheimer's Disease Stage Classification 2025 · Brain scans and imaging

More on Brain scans and imaging

• A review that covers this ground Deep Reinforcement Learning-Based Retinal Imaging in Alzheimer's Disease: Potential and Perspectives Journal of Alzheimer's disease : JAD · 2023 · usefulness score 41/100
• 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 →