Advancing Alzheimer’s Disease Detection in Clinical Settings: MRI Image Data

Dataset Overview
Dataset Name: Alzheimer’s Disease Detection Dataset

Purpose: To facilitate the development of AI and deep learning models for detecting Alzheimer’s Disease using MRI images.

Data Type: MRI brain scans in JPG format.

Labels: Four classes of Alzheimer’s Disease progression:

NonDemented: No signs of dementia.

VeryMildDemented: Very mild cognitive impairment.

MildDemented: Mild cognitive impairment.

ModerateDemented: Moderate cognitive impairment.

Dataset Size:

Train Set: 5,121 images

NonDemented: 2,560

VeryMildDemented: 1,792

MildDemented: 717

ModerateDemented: 52

Test Set: 1,279 images

NonDemented: 640

VeryMildDemented: 448

MildDemented: 179

ModerateDemented: 12

License: CC BY 4.0

Folder Structure
The dataset is organized into the following folder structure:

Alzheimer_Dataset/
│
├── images/
│ ├── train/
│ │ ├── NonDemented/
│ │ │ ├── image1.jpg
│ │ │ ├── image2.jpg
│ │ │ └── ...
│ │ ├── VeryMildDemented/
│ │ │ ├── image1.jpg
│ │ │ ├── image2.jpg
│ │ │ └── ...
│ │ ├── MildDemented/
│ │ │ ├── image1.jpg
│ │ │ ├── image2.jpg
│ │ │ └── ...
│ │ └── ModerateDemented/
│ │ ├── image1.jpg
│ │ ├── image2.jpg
│ │ └── ...
│ └── test/
│ ├── NonDemented/
│ │ ├── image1.jpg
│ │ ├── image2.jpg
│ │ └── ...
│ ├── VeryMildDemented/
│ │ ├── image1.jpg
│ │ ├── image2.jpg
│ │ └── ...
│ ├── MildDemented/
│ │ ├── image1.jpg
│ │ ├── image2.jpg
│ │ └── ...
│ └── ModerateDemented/
│ ├── image1.jpg
│ ├── image2.jpg
│ └── ...
│
└── labels/
├── train_labels.csv
└── test_labels.csv

Dataset Details
3.1 MRI Images
Description: Brain MRI scans (e.g., T1-weighted, T2-weighted) from participants.

Format: JPG

Resolution: 256x256 pixels

Source: ADNI dataset, and local hospital

Preprocessing:

Converted from DICOM to JPG.

Normalized for brightness and contrast.

Resized to a consistent resolution.

Train Labels (train_labels.csv)
The train_labels.csv file contains metadata for the training set of MRI images. Each row corresponds to an image and includes the following fields:

Image_Path: The relative path to the image file (e.g., train/NonDemented/image1.jpg).

Label: The classification label for the image, which can be NonDemented, VeryMildDemented, MildDemented, or ModerateDemented.

Participant_ID: A unique identifier for the participant (e.g., 001).

Age: The age of the participant at the time of diagnosis (e.g., 72).

Gender: The gender of the participant (e.g., Male or Female).

Diagnosis_Date: The date of diagnosis in YYYY-MM-DD format (e.g., 2022-01-15).

Notes: Additional notes or comments (e.g., N/A if no notes are available).