GMAlignNet: multi-scale lightweight brain tumor image segmentation with enhanced semantic information consistency

被引:1
作者
Song, Jianli [1 ]
Lu, Xiaoqi [1 ,2 ]
Gu, Yu [1 ]
机构
[1] Inner Mongolia Univ Sci & Technol, Sch Digital & Intelligent Ind, Inner Mongolia Key Lab Pattern Recognit & Intellig, Baotou, Peoples R China
[2] Inner Mongolia Univ Technol, Sch Informat Engn, Hohhot 010051, Peoples R China
基金
中国国家自然科学基金;
关键词
brain tumor; image segmentation; lightweight; feature alignment; U-Net; multi-scale; NET; CONVOLUTION;
D O I
10.1088/1361-6560/ad4301
中图分类号
R318 [生物医学工程];
学科分类号
0831 ;
摘要
Although the U-shaped architecture, represented by UNet, has become a major network model for brain tumor segmentation, the repeated convolution and sampling operations can easily lead to the loss of crucial information. Additionally, directly fusing features from different levels without distinction can easily result in feature misalignment, affecting segmentation accuracy. On the other hand, traditional convolutional blocks used for feature extraction cannot capture the abundant multi-scale information present in brain tumor images. This paper proposes a multi-scale feature-aligned segmentation model called GMAlignNet that fully utilizes Ghost convolution to solve these problems. Ghost hierarchical decoupled fusion unit and Ghost hierarchical decoupled unit are used instead of standard convolutions in the encoding and decoding paths. This transformation replaces the holistic learning of volume structures by traditional convolutional blocks with multi-level learning on a specific view, facilitating the acquisition of abundant multi-scale contextual information through low-cost operations. Furthermore, a feature alignment unit is proposed that can utilize semantic information flow to guide the recovery of upsampled features. It performs pixel-level semantic information correction on misaligned features due to feature fusion. The proposed method is also employed to optimize three classic networks, namely DMFNet, HDCNet, and 3D UNet, demonstrating its effectiveness in automatic brain tumor segmentation. The proposed network model was applied to the BraTS 2018 dataset, and the results indicate that the proposed GMAlignNet achieved Dice coefficients of 81.65%, 90.07%, and 85.16% for enhancing tumor, whole tumor, and tumor core segmentation, respectively. Moreover, with only 0.29 M parameters and 26.88G FLOPs, it demonstrates better potential in terms of computational efficiency and possesses the advantages of lightweight. Extensive experiments on the BraTS 2018, BraTS 2019, and BraTS 2020 datasets suggest that the proposed model exhibits better potential in handling edge details and contour recognition.
引用
收藏
页数:20
相关论文
共 59 条
[1]   CMM-Net: Contextual multi-scale multi-level network for efficient biomedical image segmentation [J].
Al-masni, Mohammed A. ;
Kim, Dong-Hyun .
SCIENTIFIC REPORTS, 2021, 11 (01)
[2]   A Review on Convolutional Neural Networks for Brain Tumor Segmentation: Methods, Datasets, Libraries, and Future Directions [J].
Balwant, M. K. .
IRBM, 2022, 43 (06) :521-537
[3]   DPAFNet: A Residual Dual-Path Attention-Fusion Convolutional Neural Network for Multimodal Brain Tumor Segmentation [J].
Chang, Yankang ;
Zheng, Zhouzhou ;
Sun, Yingwei ;
Zhao, Mengmeng ;
Lu, Yao ;
Zhang, Yan .
BIOMEDICAL SIGNAL PROCESSING AND CONTROL, 2023, 79
[4]   3D Dilated Multi-fiber Network for Real-Time Brain Tumor Segmentation in MRI [J].
Chen, Chen ;
Liu, Xiaopeng ;
Ding, Meng ;
Zheng, Junfeng ;
Li, Jiangyun .
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2019, PT III, 2019, 11766 :184-192
[5]   Info-FPN: An Informative Feature Pyramid Network for object detection in remote sensing images [J].
Chen, Silin ;
Zhao, Jiaqi ;
Zhou, Yong ;
Wang, Hanzheng ;
Yao, Rui ;
Zhang, Lixu ;
Xue, Yong .
EXPERT SYSTEMS WITH APPLICATIONS, 2023, 214
[6]  
Cicek Ozgun, 2016, Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016. 19th International Conference. Proceedings: LNCS 9901, P424, DOI 10.1007/978-3-319-46723-8_49
[7]   Survival and low-grade glioma: the emergence of genetic information [J].
Claus, Elizabeth B. ;
Walsh, Kyle M. ;
Wiencke, John K. ;
Molinaro, Annette M. ;
Wiemels, Joseph L. ;
Schildkraut, Joellen M. ;
Bondy, Melissa L. ;
Berger, Mitchel ;
Jenkins, Robert ;
Wrensch, Margaret .
NEUROSURGICAL FOCUS, 2015, 38 (01)
[8]   Effect of learning parameters on the performance of U-Net Model in segmentation of Brain tumor [J].
Das, Suchsimita ;
Swain, Mahesh ku. ;
Nayak, G. K. ;
Saxena, Sanjay ;
Satpathy, S. C. .
MULTIMEDIA TOOLS AND APPLICATIONS, 2022, 81 (24) :34717-34735
[9]   Residual 3D U-Net with Localization for Brain Tumor Segmentation [J].
Demoustier, Marc ;
Khemir, Ines ;
Nguyen, Quoc Duong ;
Martin-Gaffe, Lucien ;
Boutry, Nicolas .
BRAINLESION: GLIOMA, MULTIPLE SCLEROSIS, STROKE AND TRAUMATIC BRAIN INJURIES, BRAINLES 2021, PT I, 2022, 12962 :389-399
[10]   LFU-Net: A Lightweight U-Net with Full Skip Connections for Medical Image Segmentation [J].
Deng, Yunjiao ;
Wang, Hui ;
Hou, Yulei ;
Liang, Shunpan ;
Zeng, Daxing .
CURRENT MEDICAL IMAGING, 2023, 19 (04) :347-360