3D Brain MRI Classification for Alzheimer Diagnosis Using CNN with Data Augmentation

TL;DR

This paper presents a 3D CNN model with data augmentation that accurately classifies brain MRI scans as healthy or Alzheimer’s, achieving high accuracy and robustness, and highlights the potential of simple augmentation techniques.

Contribution

The study introduces a 3D CNN with stochastic noise injection and demonstrates its effectiveness for Alzheimer’s classification, showing significant performance improvements over resizing alone.

Findings

Test accuracy of 0.912 and AUC of 0.961

Sensitivity and specificity both exceed 0.90

Simple augmentation significantly improves classification performance

Abstract

A three-dimensional convolutional neural network was developed to classify T1-weighted brain MRI scans as healthy or Alzheimer. The network comprises 3D convolution, pooling, batch normalization, dense ReLU layers, and a sigmoid output. Using stochastic noise injection and five-fold cross-validation, the model achieved test set accuracy of 0.912 and area under the ROC curve of 0.961, an improvement of approximately 0.027 over resizing alone. Sensitivity and specificity both exceeded 0.90. These results align with prior work reporting up to 0.10 gain via synthetic augmentation. The findings demonstrate the effectiveness of simple augmentation for 3D MRI classification and motivate future exploration of advanced augmentation methods and architectures such as 3D U-Net and vision transformers.