Skip to content

Latest commit

 

History

History
35 lines (24 loc) · 2.24 KB

README.md

File metadata and controls

35 lines (24 loc) · 2.24 KB

A reproducible 3D convolutional neural network with dual attention module (3D-DAM) for Alzheimer's disease classification

The journal version of the paper describing this work is available here

Abstract

Alzheimer's disease is one of the most common types of neurodegenerative disease, characterized by the accumulation of amyloid-beta plaque and tau tangles. Recently, deep learning approaches have shown promise in Alzheimer's disease diagnosis. In this study, we propose a reproducible model that utilizes a 3D convolutional neural network with a dual attention module for Alzheimer's disease classification. We trained the model in the ADNI database and verified the generalizability of our method in two independent datasets (AIBL and OASIS1). Our method achieved state-of-the-art classification performance, with an accuracy of 91.94% for MCI progression classification and 96.30% for Alzheimer's disease classification on the ADNI dataset. Furthermore, the model demonstrated good generalizability, achieving an accuracy of 86.37% on the AIBL dataset and 83.42% on the OASIS1 dataset. These results indicate that our proposed approach has competitive performance and generalizability when compared to recent studies in the field.

Model Architecture

model architecture

Main Results

ADNI - AIBL - OASIS

Training Test Accuracy(%) Sensitivity(%) Specificity(%)
ADNI AIBL 86.3 80.2 87.1
ADNI OASIS 83.4 85.8 82.6
AIBL - OASIS ADNI 85.4 80.1 89.5

Test Performance

Citation

If you find this project useful for your research, please use the following BibTeX entries.

@misc{vu2024reproducible,
      title={A reproducible 3D convolutional neural network with dual attention module (3D-DAM) for Alzheimer's disease classification}, 
      author={Thanh Phuong Vu and Tien Nhat Nguyen and N. Minh Nhat Hoang and Gia Minh Hoang},
      year={2024},
      eprint={2310.12574},
      archivePrefix={arXiv},
      primaryClass={eess.IV}
}