Scribd
Upload
28 views4 pages

Alzheimer's Disease Classification Using Deep Learning

Uploaded by

SUMANJALI SEGU
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
28 views4 pages

Alzheimer's Disease Classification Using Deep Learning

Uploaded by

SUMANJALI SEGU
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 4

Alzheimer’s Disease Classification Using Deep

Learning on MRI Data1

Sumanjali S1, Pawan Sai Saketh1, Lahari P1, Pranav Sourya1, Palgun Kumar Reddy1, and
Gayathri R1

Amrita School of Computing, Amrita Vishwa Vidyapeetham, Bengaluru, India


{bl.en.u4cse22073, bl.en.u4cse22070, bl.en.u4cse22071, bl.en.u4cse22211, bl.en.u4cse22072,
r.gayathri}@bl.students.amrita.edu

Abstract. Alzheimer’s disease (AD) is a gradual and chronic illness of the brain
that profoundly handicaps an individual’s thinking and memory capacities. This
research uses the Alzheimer’s MRI disease classification dataset to classify
Alzheimer’s into mild demented, moderate demented, non-demented, and
very mild demented stages. We conduct the classification task with several
deep learning models, including Convolutional Neural Networks (CNN),
Residual Networks (ResNet), and Vision Transformers (ViT), and assess their
efficiency. According to the results, the models are highly effective in
classifying various types and stages of Alzheimer’s, which can be useful in
clinical practice.

Keywords: Alzheimer’s disease · MRI · deep learning · classification · convolutional


neural network · residual network · vision transformer.

1 Introduction
Alzheimer’s disease is a chronic neurodegenerative condition characterized by
progressive impairment of cognitive and memory functions. It is the most common
cause of dementia, affecting more than 50 million people globally, with projected
numbers expected to triple by 2050. Early detection of AD is crucial for timely
intervention, slowing disease progression, and improving the quality of life for
patients and their families.
This study focuses on using MRI data to classify Alzheimer’s disease stages using
deep learning models such as CNN, ResNet, and Vision Transformers. These methods
are evaluated for their efficiency and accuracy in differentiating between non-
demented, very mild demented, mild demented, and moderate demented stages of
Alzheimer’s. The study aims to provide a reliable and efficient classification system for
early diagnosis and clinical intervention.

2 Related Work
Several studies have explored deep learning methods for Alzheimer’s classification.
Korolev et al. used ResNet to classify AD from 3D MRI images, addressing
1 Supported by Amrita Vishwa Vidyapeetham.
2 S. Sumanjali et al.

the vanishing gradient problem. Basaia et al. employed CNNs to differentiate


between healthy individuals and those with Alzheimer’s, achieving high accuracy.
Recent transformer models, such as Vision Transformers, have also shown promising
results in image classification tasks, making them suitable for medical imaging.

3 Proposed Methodology
3.1 Data Preprocessing
The Alzheimer’s MRI dataset is divided into four classes: non-demented, very mild
demented, mild demented, and moderate demented. We performed data
preprocessing steps such as resizing, normalization, and data augmentation (rotation,
flipping, and zooming) to prepare the data for training.

3.2 Model Selection


We selected three deep learning models:

– CNN: A custom-designed CNN with convolutional layers, max-pooling, and fully


connected layers.
– ResNet-50: A 50-layer deep network using residual connections to improve
learning.
– Vision Transformer (ViT): Treats images as sequences of patches and uses self-
attention for classification.

3.3 Training and Evaluation


Each model was trained using cross-entropy loss and the Adam optimizer. Evaluation
metrics included accuracy, precision, recall, and F1-score. The CNN achieved 85.3%
accuracy, ResNet-50 achieved 89.6%, and ViT achieved the best accuracy at 91.2%.

4 Results
The results indicate that deep learning models are highly effective in classifying
different stages of Alzheimer’s. ViT outperformed CNN and ResNet-50, achieving the
highest accuracy. The following table presents the performance comparison of these
models.
5 Conclusion
This study demonstrates that Vision Transformers are highly effective for Alzheimer’s
stage classification using MRI data. Future work will involve testing these models in real-
world clinical settings for early diagnosis and intervention.

Alzheimer’s Disease Classification Using Deep Learning on MRI Data 3

Table 1. Model Performance Comparison


Model Accuracy Precision Recall F1-Score
CNN 85.3% 84.7% 85.1% 84.9%
ResNet-50 89.6% 89.0% 89.4% 89.2%
ViT 91.2% 90.8% 91.0% 90.9%
Acknowledgment
We thank Falah G. Salieh for providing the Alzheimer’s MRI dataset used in this study.

References
1. Alzheimer’s Association: 2019 Alzheimer’s Disease Facts and Figures. Alzheimer’s&
Dementia, 15(3), 321-387 (2019)
2. Frisoni, G. B., Fox, N. C., Jack Jr, C. R., Scheltens, P., Thompson, P. M.: Theclinical use of
structural MRI in Alzheimer disease. Nature Reviews Neurology, vol. 6, no. 2, pp. 67-77
(2010)
3. LeCun, Y., Bengio, Y., Hinton, G.: Deep learning. Nature, vol. 521, no. 7553, pp.436-444
(2015)
4. Korolev, S., Safiullin, A., Belyaev, E., Dodonova, A.: Residual and plain convolutional
neural networks for 3D brain MRI classification. In: 2017 IEEE International Symposium
on Biomedical Imaging (ISBI 2017), pp. 835-838. IEEE, Melbourne (2017)
5. Basaia, S., Agosta, N., Wagner, F., Canu, M., Magnani, E., Santangelo, S., Filippi,M.,
Horsfield, M. A.: Automated classification of Alzheimer’s disease and mild cognitive
impairment using a single MRI and deep neural networks. NeuroImage: Clinical, vol. 21,
p. 101645 (2019)
6. Islam, J., Zhang, Y.: Brain MRI analysis for Alzheimer’s disease diagnosis usingan
ensemble system of deep convolutional and recurrent neural networks. Brain
Informatics, vol. 5, pp. 2-12 (2018)
7. Dosovitskiy, A., Beyer, L., Kolesnikov, A., et al.: An Image is Worth 16x16 Words:
Transformers for Image Recognition at Scale. arXiv preprint arXiv:2010.11929 (2020)
8. Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., Guo, B.: Swin Transformer:
Hierarchical Vision Transformer using Shifted Windows. In: Proceedings of the IEEE/CVF
International Conference on Computer Vision (ICCV), pp. 1001210022 (2021)
9. Zhang, D., Wang, Y., Zhou, L., Yuan, H., Shen, D.: Multimodal classification ofAlzheimer’s
disease and mild cognitive impairment. NeuroImage, vol. 55, no. 3, pp. 856-867 (2011)
10. He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition.In: 2016
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 770-778. IEEE,
Las Vegas (2016)
11. Dimitriadis, S., Panagopoulos, I., Angelopoulos, P., et al.: Alzheimer’s disease
classification using MRI images: A comprehensive review. In: 2019 IEEE International
Conference on Imaging Systems and Techniques (IST), pp. 1-6. IEEE, Thessaloniki (2019)
4 S. Sumanjali et al.

12. Cao, Y., Li, Q., Li, Y., et al.: A review on convolutional neural networks in medicalimage
analysis. IEEE Access, vol. 7, pp. 103049-103066 (2019)
13. Ota, N., Matsubara, K., Hara, H., et al.: A method for classifying Alzheimer’sdisease with
functional connectivity using deep learning. In: 2019 41st Annual International Conference
of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 186-189. IEEE, Berlin
(2019)
14. Huang, H., Lin, X., Jiang, X., et al.: Diagnosis of Alzheimer’s disease using deeplearning-
based MR image analysis. In: 2019 IEEE International Conference on Bioinformatics and
Biomedicine (BIBM), pp. 297-302. IEEE, San Diego (2019)
15. Sani, R. F., Stojanovic, J., Tomovic, R., et al.: Alzheimer’s disease diagnosis usingdeep
learning techniques: A systematic review. Information Sciences, vol. 527, pp. 1-21 (2020)

You might also like