Augmented Alzheimer MRI Dataset

MRI image dataset for the classification of stages of Alzheimer's dementia, including original images and augmented versions to increase data diversity.

Download dataset

Size

40,400 MRI images in JPEG/PNG format, with original and augmented data

Licence

GNU Lesser General Public License (LGPL)

Description

‍

The Augmented Alzheimer MRI Dataset contains more than 40,000 MRI images divided into four classes corresponding to the stages of dementia: Non-demented, Mild, Moderate, Very mild. The dataset includes original images as well as augmented versions to enrich the training base.

‍

What is this dataset for?

‍

Training computer vision models for the detection and classification of stages of Alzheimer's disease
Improving medical diagnostic tools using MRI imaging
Enabling neuroimaging and deep learning research on augmented data

‍

Can it be enriched or improved?

‍

The dataset can be supplemented by additional clinical annotations, the integration of anonymized patient metadata, or the addition of new augmentation methods to increase the robustness of the models. Normalizing and preprocessing images can also improve performance.

‍

🔎 In summary

Criterion	Evaluation
🧩 Ease of use	⭐⭐⭐✩✩ (Requires knowledge in medical imaging)
🧼 Need for cleaning	⭐⭐⭐✩✩ (Moderate – check for duplicates and image consistency)
🏷️ Annotation richness	⭐⭐⭐✩✩ (Average – well-defined classes but few associated metadata)
📜 Commercial license	⚡ Yes (LGPL, commercial use possible)
👨‍💻 Beginner friendly	⚠️ Medium – useful with guidance in medical imaging
🔁 Fine-tuning ready	🧠 Suitable for classification and detection in neuroimaging
🌍 Cultural diversity	⚠️ Not specified, medical data of unknown origin