Enhancing Medical Image Classification With Context Modulated Attention and Multi-Scale Feature Fusion

被引:1
作者
Zhang, Renhan [1 ,2 ]
Luo, Xuegang [3 ]
Lv, Junrui [3 ]
Cao, Junyang [1 ,2 ]
Zhu, Yangping [4 ]
Wang, Juan [1 ,2 ]
Zheng, Bochuan [1 ,2 ]
机构
[1] China West Normal Univ, Sch Comp Sci, Nanchong 637009, Sichuan, Peoples R China
[2] China West Normal Univ, Inst Artificial Intelligence, Nanchong 637000, Sichuan, Peoples R China
[3] Panzhihua Univ, Sch Math & Comp Sci, Panzhihua 617000, Sichuan, Peoples R China
[4] Nanjiang TCM Hosp, Dept Radiol, Bazhong 636600, Sichuan, Peoples R China
来源
IEEE ACCESS | 2025年 / 13卷
关键词
Medical diagnostic imaging; Transformers; Feature extraction; Image classification; Semantics; Context modeling; Convolutional neural networks; Modulation; Computational modeling; Mathematical models; Medical images; global semantics; local features; transformer; context modulated attention; multi-stage feature fusion network; lesion recognition; performance indicators; feature extraction; semantic feature representation;
D O I
10.1109/ACCESS.2025.3532354
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
This research proposes a multi-stage feature fusion network (MSFF) for medical image classification. In view of the problems existing in medical images, such as noise, diversity, and similarity among different classes, MSFF enhances the global context perception in the window partitioning framework through Context Modulation Attention (CMA). Meanwhile, it extracts fine-grained local information via the multi-stage Contextual Information Refinement (CIR) module and gradually fuses multi-stage local and global features to generate richer semantic representations. The experimental results demonstrate that MSFF significantly outperforms existing methods in multiple performance metrics (including accuracy, precision, recall, F1-score, Matthews Correlation Coefficient (MCC), Kappa coefficient, Area Under the Curve (AUC), balanced accuracy, and geometric mean) on four datasets (Endoscopic Bladder Tissue, Kvasir, SARS-COV-2 Ct-Scan, and Thyroid Nodule), showing its excellent performance in the task of medical image classification.
引用
收藏
页码:15226 / 15243
页数:18
相关论文
共 55 条
[1]  
Alexey D, 2020, arXiv, DOI [arXiv:2010.11929, DOI 10.48550/ARXIV.2010.11929]
[2]   A Novel Method for COVID-19 Diagnosis Using Artificial Intelligence in Chest X-ray Images [J].
Almalki, Yassir Edrees ;
Qayyum, Abdul ;
Irfan, Muhammad ;
Haider, Noman ;
Glowacz, Adam ;
Alshehri, Fahad Mohammed ;
Alduraibi, Sharifa K. ;
Alshamrani, Khalaf ;
Basha, Mohammad Abd Alkhalik ;
Alduraibi, Alaa ;
Saeed, M. K. ;
Rahman, Saifur .
HEALTHCARE, 2021, 9 (05)
[3]   Magnetic resonance imaging-based brain tumor grades classification and grading via convolutional neural networks and genetic algorithms [J].
Anaraki, Amin Kabir ;
Ayati, Moosa ;
Kazemi, Foad .
BIOCYBERNETICS AND BIOMEDICAL ENGINEERING, 2019, 39 (01) :63-74
[4]  
Angelov P, 2020, medRxiv, DOI [10.1101/2020.04.24.20078584, 10.1101/2020.04.24.20078584, DOI 10.1101/2020.04.24.20078584, DOI 10.1101/2020.04.24.20078584V3]
[5]  
Angelov P, 2020, Explainable-by-design approach for Covid-19 classification via CTscan, DOI [10.1101/2020.04.24.20078584, DOI 10.1101/2020.04.24.20078584]
[6]  
Anqi M., 2023, P INT C MACH LEARN, P1
[7]   TS-CNN: A Three-Tier Self-Interpretable CNN for Multi-Region Medical Image Classification [J].
Ashwath, V. A. ;
Sikha, O. K. ;
Benitez, Raul .
IEEE ACCESS, 2023, 11 :78402-78418
[8]  
Ba J.L., 2016, arXiv
[9]  
Carion Nicolas, 2020, End-toend object detection with transformers
[10]   GasHis-Transformer: A multi-scale visual transformer approach for gastric histopathological image detection [J].
Chen, Haoyuan ;
Li, Chen ;
Wang, Ge ;
Li, Xiaoyan ;
Rahaman, Md Mamunur ;
Sun, Hongzan ;
Hu, Weiming ;
Li, Yixin ;
Liu, Wanli ;
Sun, Changhao ;
Ai, Shiliang ;
Grzegorzek, Marcin .
PATTERN RECOGNITION, 2022, 130