文献类型： 外文期刊
作者： Ge, Yanning 1 ; Xu, Li 1 ; Wang, Xiaoding 1 ; Que, Youxiong 2 ; Piran, Md. Jalil 4 ;
作者机构： 1.Fujian Normal Univ, Coll Comp & Cyber Secur, Fujian Prov Key Lab Network Secur & Cryptol, Fuzhou 350117, Peoples R China
2.Chinese Acad Trop Agr Sci, Inst Trop Biosci & Biotechnol, Natl Key Lab Trop Crop Breeding, Sanya Res Inst, Haikou 571101, Peoples R China
3.Fujian Agr & Forestry Univ, Natl Engn Res Ctr Sugarcane, Key Lab Sugarcane Biol & Genet Breeding, Minist Agr & Rural Affairs,Coll Agr, Fuzhou 350002, Peoples R China
4.Sejong Univ, Dept Comp Sci & Engn, Seoul 05006, South Korea
关键词： Tumors; Magnetic resonance imaging; Feature extraction; Computed tomography; Accuracy; Brain modeling; Medical diagnostic imaging; Image recognition; Clustering algorithms; Convolutional neural networks; Brain tumor detection; multimodal image processing; contrastive learning; dual-branch network; adaptive weight mask
期刊名称：IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS （ 影响因子：6.8； 五年影响因子：7.0 ）
ISSN： 2168-2194
年卷期： 2025 年 29 卷 8 期
页码：
收录情况： SCI
摘要： Brain tumor detection has advanced significantly with the development of deep learning technology. Although multimodal data, such as Magnetic Resonance Imaging (MRI) and Computed Tomography (CT), has potential advantages in diagnostics, most existing studies rely solely on a single modality. This is because common fusion methods may lead to the loss of critical information when attempting multimodal fusion. Therefore, effectively integrating multimodal data has become a significant challenge. Additionally, medical image analysis requires large amounts of annotated data, and labeling images is a resource-intensive task that demands experienced professionals to spend a considerable amount of time. To address these challenges, this paper introduces a new unsupervised learning framework named Double-SimCLR. This framework builds on the foundation of contrastive learning and features a dual-branch structure, enabling direct and simultaneous processing of MRI and CT images for multimodal feature fusion. Given the "weak feature" characteristics of CT images (e.g., low soft tissue contrast and low resolution), we incorporated adaptive weight masking technology to enhance CT feature extraction. Moreover, we introduced a multimodal attention mechanism, which ensures that the model focuses on salient information, thereby elevating the precision and robustness of brain tumor detection. Even without substantial labeled data, experimental results demonstrate that Double-SimCLR achieves 93.458% accuracy, 92.463% precision, and a 93.058% F1-score, outperforming state-of-the-art (SOTA) models by 2.871%, 2.643%, and 3.098%, respectively.
- 相关文献
作者其他论文 更多>>
-
Sugarcane Varieties GT 42 and GT 44, not a Flash in the Pan, but the Flag in the Ship
作者：Wu, Qibin;Li, Zhenxiang;Li, Aomei;Zhang, Yuebin;Que, Youxiong;Wu, Qibin;Zhang, Yuebin;Que, Youxiong;Li, Zhenxiang;Que, Youxiong;Li, Aomei;Wang, Lunwang;Tang, Shiyun;Zhou, Hui
关键词：Sugarcane; GT 42 and GT 44; Breeding process; Characteristics; Biological basis
-
Blockchain-Empowered H-CPS Architecture for Smart Agriculture
作者：Wang, Xiaoding;Wu, Qibin;Que, Youxiong;Wang, Xiaoding;Zeng, Haitao;Yang, Xu;Cui, Hui;Yi, Xun;Piran, Md. Jalil;Luo, Ming;Que, Youxiong
关键词：blockchain; human-cyber-physical systems; prospect; semantic blockchain; smart agriculture
-
FCDSB: A Fog Computing Network Architecture Based on Dynamic Sharding Blockchain for Consumer Electronics in AIoT
作者：Wei, Chenxiang;Lin, Hui;Wang, Xiaoding;Wei, Chenxiang;Lin, Hui;Que, Youxiong;Wang, Xiaoding;Que, Youxiong
关键词：Artificial Intelligence of Things; consumer elec-tronics; fog computing; sharding blockchain; sharding blockchain; deep reinforcement learning; deep reinforcement learning; deep reinforcement learning
-
Multi-omics profiling of Nanhaia speciosa and Nanhaia fordii: insights into lectin dynamics, nodulation, and triterpenoid saponin biosynthesis
作者：Li, Zhiying;Wang, Jiabin;Fu, Yunliu;Jing, Yonglin;Huang, Bilan;Meng, Chunyang;Xu, Li;Li, Zhiying;Wang, Jiabin;Fu, Yunliu;Jing, Yonglin;Huang, Bilan;Meng, Chunyang;Xu, Li;Li, Zhiying;Wang, Jiabin;Fu, Yunliu;Jing, Yonglin;Huang, Bilan;Meng, Chunyang;Xu, Li;Li, Zhiying;Wang, Jiabin;Fu, Yunliu;Jing, Yonglin;Huang, Bilan;Meng, Chunyang;Xu, Li;Zhang, Junfang;Wu, Huazhou
关键词：
Nanhaia speciosa ;Nanhaia fordii ; Chromosome-level genome; Licorice saponins; Isoflavonoids; Nodulation; Lectin -
Structure and changing trends of carbon footprint in sugarcane
作者：Xie, Yuan;He, Ting;Que, Yuyue;Chen, Rukai;Que, Youxiong;Wu, Qibin;Que, Youxiong;Que, Yuyue;Chen, Niandong
关键词：sugarcane carbon footprint; structural analysis; sensitivity analysis; individual fixed regression
-
Identification and Characterization of Key Genes for Nitrogen Utilization from Saccharum spontaneum Sub-Genome in Modern Sugarcane Cultivar
作者：Hui, Qianlong;Song, Ting;Yang, Dantong;Wu, Qibin;Guo, Jinlong;Que, Youxiong;Xu, Liping;Wu, Qibin;Que, Youxiong
关键词：sugarcane; nitrogen use efficiency; genetic mechanisms; sub-genome;
ScNRT2.3 -
International research initiative on genomics-guided sugarcane breeding
作者：Zhang, Xingtan;Wang, Jungang;Que, Youxiong;Lin, Jishan;Zhang, Xingtan;Noor-ul-Ain;Abid, Javeria;Wang, Ren;Zhang, Jisen;Bao, Yixue;Zhang, Muqing;Zhang, Jisen;Bao, Yixue;Zhang, Muqing;Zhao, Peifang;Zhang, Yuebin;Chen, Li-yu;Ming, Ray;Herry, Robert;Tangphatsornruang, Sithichoke;Wang, Qinnan;Wang, Qinnan;Huang, Dongliang;Ye, Jian;Luo, Ming;Butterfield, Mike
关键词：
