一种基于CBAM和改进通道注意力的EfficientNet的混合脑肿瘤分类方法

摘要/Abstract

摘要： 为进一步提高脑肿瘤影像诊断的准确性和稳健性，提出一种基于CBAM（Convolutional Block Attention Module）和改进通道注意力机制的EfficientNet神经网络（IC+IEffxNet）的新型混合脑肿瘤分类方法。该方法分为2个阶段，第一阶段由基于改进空间注意力机制的CBAM模型提取特征。第二阶段将EfficientNet架构中的Squeeze and Excitation（SE）块替换成Efficient Channel Attention （ECA）块，将第一阶段的组合特征输出作为第二阶段的输入。实验展示了在混合脑肿瘤MRI数据集下，神经胶质瘤患者、脑膜瘤患者、脑垂体瘤患者与正常患者图像的4分类结果，实验结果显示分类平均准确率比现有方法提高约0.5~2个百分点。实验结果证明了该方法的有效性，为医疗专家能够准确判断脑肿瘤种类提供了新的参考。

关键词: 脑肿瘤, ECA, EfficientNet, CBAM, 分类

Abstract: In order to further improve the accuracy and robustness of brain tumor image diagnosis， a novel hybrid brain tumor classification method based on CBAM（Convolutional Block Attention Module） and EfficientNet with improved channel attention mechanism （IC+IEffxNet） is proposed. The method is divided into 2 stages. In the first stage， the features will be extracted by CBAM model based on improved spatial attention mechanism. In the second stage， the sequence and exception （SE） block in EfficientNet architecture is replaced by the efficient channel attention （ECA） block， and the combined feature output of the first stage is used as the input of the second stage. Experiment shows the 4 classifications of glioma， meningioma， pituitary and normal images from the mixed brain tumor MRI dataset. The results show that the average classification accuracy is about 0.5~2 percentage points higher than the existing methods. The experimental results demonstrate the effectiveness of the method and provide a new reference for medical experts to accurately judge brain tumor.

Key words: brain tumor, ECA, EfficientNet, CBAM, classification

华昕宇, 祁云嵩. 一种基于CBAM和改进通道注意力的EfficientNet的混合脑肿瘤分类方法[J]. 计算机与现代化, 2023, 0(05): 1-7.

HUA Xin-yu, QI Yun-song. A Hybrid Brain Tumor Classfication Study Based on CBAM and EfficientNet with Improved Channel Attention[J]. Computer and Modernization, 2023, 0(05): 1-7.

参考文献［27］

［1］	刘梦涵，张光. 基于深度学习的脑肿瘤影像研究进展［J］.影像技术， 2021，33（5）:50-55.
［2］	ZHOU Y F， LI Z S，ZHU H，et al. Holistic brain tumor screening and classification based on densenet and recurrent neural network［C］// International MICCAI Brainlesion Workshop. 2018:208-217
［3］	KALAISELVI T， PADMAPRIYA S T， SRIRAMAKRISHN
	AN P， et al. Deriving tumor detection models using convolutional neural networks from MRI of human brain scans［J］. International Journal of Information Technology， 2020，12:403-408.
［4］	TO[G]GA[O]AR M，C[O]MERT Z，ERGEN B. Classification of brain MRI using hyper column technique with convolutional neural network and feature selection method［J］. Expert Systems with Applications， 2020，149:113274.
［5］	TO[G]GA[O]AR M， ERGEN B， C[O]MERT Z. BrainMRNet: brain tumor detection using magnetic resonance images with a novel convolutional neural networkmodel［J］. Medical Hypotheses，2020，134:109531.
［6］	GHASSEMI N， SHOEIBI A， ROUHANI M. Biomedical signal processing and control deep neural network with generative adversarial networks pre-training for brain tumor classification based on MR images［J］. Biomedical Signal Processing and Control， 2020，57:101678.
［7］	BEGUM S S， LAKSHMI D R. Combining optimal wavelet statistical texture and recurrent neural network for tumour detection and classification over MRI［J］. Multimedia Tools and Applications，2020，79（19-20）:14009-14030.
［8］	DESHPANDE A， ESTRELA V V， PATAVARDHAN P. The DCT-CNN-ResNet50 architecture to classify brain tumors with super-resolution， convolutional neural network， and the ResNet50［J］. Neuroscience Informatics， 2021，1（4）:100013.
［9］	GE C， GU I H， JAKOLA A S， et al. Enlarged training dataset by pairwise GANs for molecular-based brain tumor classification［J］. IEEE Access， 2020，8:22560-22570.
［10］	HUANG Z G， DU X H， CHEN L M， et al. Convolutional neural network based on complex networks for brain tumor image classification with a modified activation function［J］. IEEE Access， 2020，8:89281-89290.
［11］	TAN M， LE Q V. EfficientNet: Rethinking model scaling for convolutional neural networks［J］. arXiv preprint arXiv:1905.11946v5， 2019.
［12］	WOO S， PARK J， LEE J Y， CBAM: Convolutional block attention module［C］// Proceedings of the European Conference on Computer Vision（ECCV）. 2018:3-19.
［13］	HE K M， ZHANG X Y， REN S Q， et al. Deep eesidual learning for image recognition［C］// IEEE Conference on Computer Vision and Pattern Recognition. 2016:770-778.
［14］	WANG Q L， WU B G， P. ZHU P F， et al. ECA-net: Efficient channel attention for deep convolutional neural networks［C］// 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition（CVPR）. 2020.DOI:10.1109/CVPR42600.2020.01155.
［15］	ZHU M， MIN W D， WANG Q， et al. PFLU and FPFLU: Two novel non-monotonic activation functions inconvolutional neural networks［J］. Neurocomputing， 2021，429:110-117.
［16］	PANDIAN A A， BALASUBRAMANIAN R. Fusion of contourlet transform and zernike moments using content based image retrieval for MRI brain tumor images［C］. Indian Journal of Science and Technology， 2016，9 （29）. DOI:10.17485/ijst/2016/v9i29/93837.
［17］	ANTONY A， FATHIMA K A， RAJU D， et al. Brain tumor detection and classification in MRI images［J］. IEEE Acess，2017，6:84-89.
［18］	SEETHA J， RAJA S S. Brain tumor classification using convolutional neural networks. Biomed［J］. Journal de Pharmacology （Paris）， 2018，11（3）:1457-1461.
［19］	SULTAN H H， SALEM N M， AL-ATABANY W. Multi-classification of brain tumor images using deep neural network［J］. IEEE Access，2019，7:69215-69225.
［20］	UCUZAL H， YASAR S， COLAK C. Classification of brain tumor types by deep learning with convolutional neural network on magnetic resonance images using a developed web-based interface［C］// The 3rd International Symposium on Multidisciplinary Studies and Innovative Technologies（ISMSIT）. 2019. DOI: 10.1109/ISMSIT.2019.8932761.
［21］	冷浩柏，卢涵宇，郭彩，等. 基于改进胶囊神经网络的遥感影像分类［J］. 计算机与现代化， 2022（2）:102-107.
［22］	WU J Q， LI G X， LU H M， et al. A supervoxel classification based method for multi-organ segmentation from abdominal CT images［J］. Journal of Image and Graphics， 2021，9（1）:9-14.
［23］	WU Y J， TSAI C M， SHIH F. Improving leaf classification rate via background removal and ROI extraction［J］. Journal of Image and Graphics， 2016，4（2）:93-98.
［24］	KARAYANEVA Y， HINTEA D. Object recognition in Python and MNIST dataset modification and recognition with five machine learning classifiers［J］. Journal of Image and Graphics， 2018，6（1）:10-20.
［25］	ONO K， IWAMOTO Y， CHEN Y W， et al. Automatic segmentation of infant brain ventricles with hydrocephalus in MRI based on 2.5D U-Net and transfer learning［J］. Journal of Image and Graphics， 2020，8（2）:42-46.
［26］	DOVGANICH A A， KHVOSTIKOV A V， PCHELINTSEV Y A， et al. Automatic out-of-distribution detection methods for improving the deep learning classification of pulmonary X-ray images［J］. Journal of Image and Graphics， 2022，10（2）:56-63.
［27］	姚传文，黄道斌，叶明全. 基于粗糙集自适应粒度的MR脑肿瘤图像分割［J］. 计算机与现代化， 2021（1）:34-37.

[1]	孟娜, 方维维, 路红英. 一种面向微控制器上环境声音分类的DNN压缩方法[J]. 计算机与现代化, 2024, 0(01): 80-86.
[2]	邱凯星, 冯广. 基于双重特征注意力的多标签图像分类模型[J]. 计算机与现代化, 2023, 0(12): 41-47.
[3]	唐诗琪, 周瑞平, 谢仕斌, 刘梦赤, 肖文, . 基于栈式降噪编码器的跨语言多标签情感分类[J]. 计算机与现代化, 2023, 0(11): 6-12.
[4]	吴甜, 刘海华, 童顺延. 基于深度反馈的卷积神经网络的图像分类[J]. 计算机与现代化, 2023, 0(09): 82-86.
[5]	刘禅奕, 黄丹, 薛林雁, 王涛, 朱桃, . 改进EfficientNet网络的COVID-19 X光分类[J]. 计算机与现代化, 2023, 0(09): 94-99.
[6]	马国祥, 杨凌菲, 严传波, 张志豪, 孙彬, 王晓荣. 基于深度DenseNet网络的肝包虫病超声影像诊断方法[J]. 计算机与现代化, 2023, 0(09): 100-104.
[7]	欧阳飞, 吴旭, 向东升. 基于改进YOLOX的垃圾分类检测方法[J]. 计算机与现代化, 2023, 0(08): 68-73.
[8]	李兰兰, 高建龙, 朱晓, 穆培政. 基于区域内读数段分类的插入/缺失基因组变异检测方法[J]. 计算机与现代化, 2023, 0(07): 13-19.
[9]	秦竹媛, 吴浩忠, 谭代庆, 韩爱庆, 臧昊, 王选, 唐燕. 基于多尺度ResNet融合注意力机制的麦冬细粒度识别[J]. 计算机与现代化, 2023, 0(07): 105-111.
[10]	朱剑波, 葛明锋, 董文飞. 基于改进EfficientNet的阿尔兹海默症图像分类[J]. 计算机与现代化, 2023, 0(06): 56-61.
[11]	王杰, 潘凤, 张艳莎, 谭棉, 严晓波, 王林, . 融合带权非局部模块的铝型材表面缺陷分类[J]. 计算机与现代化, 2023, 0(05): 86-92.
[12]	徐涯昕, 何泽恩, 徐绪堪. 基于CNN-BiLSTM网络的数控机床故障文本自动分类[J]. 计算机与现代化, 2023, 0(04): 7-14.
[13]	魏鑫, 何小海, 滕奇志, 卿粼波, 陈洪刚. 基于BERT-BiLSTM-Attention混合模型的事件抽取方法[J]. 计算机与现代化, 2023, 0(04): 26-31.
[14]	朱原冶, 倪建军, 唐广翼. 一种基于改进卷积神经网络的RGB-D室内场景分类方法[J]. 计算机与现代化, 2023, 0(04): 73-77.
[15]	马瑜涓, 韩建宁, 史韶杰, 曹尚斌, 杨志秀. 基于HMRF的改进Kmeans脑肿瘤分割算法[J]. 计算机与现代化, 2023, 0(03): 1-5.