A Study on Improved Swin Transformer for Alzheimers Disease MRI Image Classification

Corresponding author: Wang yanli, wan.yanli@imicams.ac.cn

DOI: 10.12201/bmr.202505.00029

•  

  Abstract: Purpose/Significance: Early diagnosis of Alzheimers disease (AD) is crucial for improving patients quality of life and treatment outcomes. With the advancement of medical imaging technology, MRI images have become an important tool for identifying and diagnosing AD. However, improving the accuracy of MRI image classification remains a challenge. Methods: To address this issue, this study proposes an improved Swin Transformer model, combined with a Multi-Resolution Feature Fusion (MRFF) module. The method aims to enhance the models ability to recognize features at different scales, thereby improving classification accuracy. Results: Experimental results show that the improved Swin Transformer + MRFF model significantly outperforms the baseline model on the OASIS1 dataset. The model achieves a classification accuracy of 87.26% and a recall rate of 91.43%. Compared to the baseline model, the improved models accuracy increased by 2.43 percentage points, and the F1 score rose from 87.26% to 91.42%. Additionally, the MRFF module effectively captures both local details and global structures of the images, significantly improving the recognition of Mild Demented and Non-Demented categories. By incorporating data augmentation and expansion strategies, the issues of insufficient sample size and class imbalance were addressed, further enhancing the models performance. Conclusion: The results demonstrate that the improved deep learning method based on Swin Transformer has excellent potential for Alzheimers MRI image classification tasks. This approach provides strong support for early diagnosis and can effectively improve the diagnostic accuracy of Alzheimers disease.

  Key words: Swin Transformer; Multi-Resolution Feature Fusion (MRFF); MR Medical Image Classification; Alzheimers Disease.

  Submit time: 21 May 2025

  Copyright: The copyright holder for this preprint is the author/funder, who has granted biomedRxiv a license to display the preprint in perpetuity.
• html

• 图表

• The guiding significance of quantitative EEG parameters in adjusting treatment plans for Alzheimers disease. 2025. doi: 10.12201/bmr.202501.00062

  SUI XIN, SONG JUAN, BAO YANAN, LIN YU, MAN ZITENG, CHENG TIANTIAN, YANG HONGYAN. 1. School of Pharmacy,Qiqihar Medical University,Qiqihar 161006,China2. Affiliated Third Hospital,Qiqihar Medical University,Qiqihar 161006,China. 2024. doi: 10.12201/bmr.202407.00036

  XIANG Hongying, LING Yingchun. Analysis of the predictive value of quantitative parameters of electroencephalogram and serum miR-146a-5p for cognitive impairment in Alzheimers disease patients. 2024. doi: 10.12201/bmr.202412.00027

  gongxiaocui, anxinying. Research on Semantic Feature Enhancement for Medical Literature Classification. 2024. doi: 10.12201/bmr.202411.00088

  刘, Yan Zhu, Zongyou Li, Dongfei Lin, Lihong Liu, Dongyun Shi. A study on Diseases Classification and Modelof the SNOMED CT. 2021. doi: 10.12201/bmr.202110.00005

  张胜发, zhou wei. A Research on the Classification of Biomedical big Data for Open Applications. 2024. doi: 10.12201/bmr.202411.00082

  wangying. The teaching design and implementation of online + offline digital image processing course with OBE concept. 2021. doi: 10.12201/bmr.202108.00011

  zhaocongpu, YUAN Da, ZHU Pu-jue, ZHOU Jiong, CHEN Zheng, PENG Hua. Research and practice on intelligent classification of medical safety incidents based on deep BERT. 2023. doi: 10.12201/bmr.202312.00021

  Duan Lei, Zhang Xiang, Zhang Yan. Study on the Measurement and Application of Diagnostic and Treatment Difficulty in Disease Categories. 2024. doi: 10.12201/bmr.202403.00007

  zhang chen, chen hui, cao feng, wang yueqi, ke ren. Research progress of automatic segmentation of left atrial CTA image based on deep learning. 2025. doi: 10.12201/bmr.202503.00042

• ID	Submit time	Number	Download
  1	2025-02-06	bmr.202505.00029V1	Download

• Public  Anonymous  To author only Submit Comments