Share：
Share this content in WeChat
X
[Chinese][PDF]5336
• Technical Article •
MRI brain tumor classification based on multi-scale residual network
HUANG Min  XIONG Zhengyun  ZHU Junlin 

Cite this article as: HUANG M , XIONG Z Y, ZHU J L. MRI brain tumor classification based on multi-scale residual network[J]. Chin J Magn Reson Imaging, 2023, 14(1): 124-129. DOI:10.12015/issn.1674-8034.2023.01.022.


[Abstract] Objective Research and build artificial intelligence deep learning network to achieve high accuracy MRI brain tumor four classification on two public brain MR image datasets.Materials and Methods We propose a multi-scale residual network for MRI brain tumor classification model to achieve the task of four brain tumor classification. The model consists of four modules: multi-scale input, improved residuals, down-sampling, and dual-channel pooling. Brain MR images from normal subjects in Kaggle and tumor patients in Figshare are combined to train and evaluate the performance of the proposed model.Results The model is tested on 352 MR images. When only multi-scale input module is used, the average classification accuracy is 96.59%. After adding the subsampling module, the accuracy reaches 98.58%. Compare max-pooling, mean-pooling and dual-channel pooling, the accuracy is 96.02%, 97.16% and 98.58%, respectively. The multi-scale residual network has a good classification effect on brain tumors, and the classification accuracy of glioma, meningioma, pituitary tumor and normal images is 99.14%, 99.14%, 99.42% and 99.42%, respectively.Conclusions MRI is a typical medical imaging method for the examination of brain tumors, but the accurate classification of brain tumors manually by radiologists is extremely subjective and uncertain. The proposed multi-scale residual network can provide an effective method for automatic classification of brain tumors, and it can improve the accuracy of MRI brain tumor classification. It solves the problem of gradient vanishing well and improves the generalization ability of the model.
[Keywords] artificial intelligence；deep learning；magnetic resonance imaging；brain tumor classification；multi-scale residual network；down-sampling；dual-channel pooling；convolutional neural network

HUANG Min1, 2*   XIONG Zhengyun1   ZHU Junlin1  

1 School of Biomedical Engineering, South-Central Minzu University, Wuhan 430074, China

2 Hubei Key Laboratory of Medical Information Analysis and Tumor Diagnosis & Treatment, Wuhan 430074, China

Corresponding author: Huang M, E-mail: minhuang@mail.scuec.edu.cn

Conflicts of interest   None.

ACKNOWLEDGMENTS Natural Science Foundation of Hubei Province (No. 2020CFB837).
Received  2022-08-12
Accepted  2022-11-14
DOI: 10.12015/issn.1674-8034.2023.01.022
Cite this article as: HUANG M , XIONG Z Y, ZHU J L. MRI brain tumor classification based on multi-scale residual network[J]. Chin J Magn Reson Imaging, 2023, 14(1): 124-129. DOI:10.12015/issn.1674-8034.2023.01.022.

[1]
BIDKAR P S, KUMAR R, GHOSH A. SegNet and Salp Water Optimization-driven Deep Belief Network for Segmentation and Classification of Brain Tumor[J/OL]. Gene Expr Patterns, 2022, 45, 119248 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S1567133X22000187. DOI: 10.1016/j.gep.2022.119248.
[2]
ISMAEL S A A, MOHAMMED A, HEFNY H. An enhanced deep learning approach for brain cancer MRI images classification using residual networks[J/OL]. Artif Intell Med, 2020, 102, 101779 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S0933365719306177. DOI: 10.1016/j.artmed.2019.101779.
[3]
VANKDOTHU R, HAMEED M A, FATIMA H. A Brain Tumor Identification and Classification Using Deep Learning based on CNN-LSTM Method[J/OL]. Comput Electr Eng, 2022, 101: 107960 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S0045790622002361. DOI: 10.1016/j.compeleceng.2022.107960.
[4]
NASER M A, DEEN M J. Brain tumor segmentation and grading of lower-grade glioma using deep learning in MRI images[J/OL]. Comput Biol Med, 2020, 121: 103758 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S0010482520301335. DOI: 10.1016/j.compbiomed.2020.103758.
[5]
RAJEEV S K, RAJASEKARAN M P, VISHNUVARTHANAN G, et al. A biologically-inspired hybrid deep learning approach for brain tumor classification from magnetic resonance imaging using improved gabor wavelet transform and Elmann-BiLSTM network[J/OL]. Biomed Signal Proces, 2022, 78: 103949 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S1746809422004487. DOI: 10.1016/j.bspc.2022.103949.
[6]
LU Z, BAI Y, CHEN Y, et al. The Classification of Gliomas Based on a Pyramid Dilated Convolution ResNet Model[J]. Pattern Recongn Lett, 2020, 133(5): 173-179. DOI: 10.1016/j.patrec.2020.03.007.
[7]
KUMAR R L, KAKARLA J, ISUNURI B V, et al. Multi-class brain tumor classification using residual network and global average pooling[J]. Multimed Tools and Appl, 2021, 80(9): 13429-13438. DOI: 10.1007/s11042-020-10335-4.
[8]
FRANCISCO JAVIER DÍAZ-PERNAS, MARTÍNEZ-ZARZUELA MARIO, ANTÓN-RODRÍGUEZ MÍRIAM, et al. A Deep Learning Approach for Brain Tumor Classification and Segmentation Using a Multiscale Convolutional Neural Network[J]. Healthcare, 2021, 9(2): 153-166. DOI: 10.3390/healthcare9020153.
[9]
SOBHANINIA Z, KARIMI N, KHADIVI P, et al. Brain Tumor Classification by Cascaded Multiscale Multitask Learning Framework Based on Feature Aggregation[J/OL]. arXiv e-prints, 2021 [2022-08-10]. https://arxiv.org/ftp/arxiv/papers/2112/2112.14320.pdf. DOI: 10.48550/arXiv.2112.14320.
[10]
HAZARIKA RA, ABRAHAM A, KANDAR D, et al. An Improved LeNet-Deep Neural Network Model for Alzheimer's Disease Classification Using Brain Magnetic Resonance Images[J]. IEEE Access, 2021, 9: 161194-161207. DOI: 10.1109/ACCESS.2021.3131741.
[11]
CHANG J, ZHANG L, GU N, et al. A mix-pooling CNN architecture with FCRF for brain tumor segmentation[J]. J Vis Commun Image R, 2019, 58: 316-322. DOI: 10.1016/j.jvcir.2018.11.047.
[12]
WANG K, JIANG P, MENG J, et al. Attention-Based DenseNet for Pneumonia Classification[J]. IRBM, 2022, 43(5): 479-485. DOI: 10.1016/j.irbm.2021.12.004.
[13]
ALNOWAMI M, TAHA E, ALSEBAEAI S, et al. MR image normalization dilemma and the accuracy of brain tumor classification model[J]. J Radiat Res Appl Sc, 2022, 15(3): 33-39. DOI: 10.1016/j.jrras.2022.05.014.
[14]
SUNSUHI G S, JOSE S A. An Adaptive Eroded Deep Convolutional neural network for brain image segmentation and classification using Inception ResnetV2[J/OL]. Biomed Signal Proces, 2022, 78: 103863 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S1746809422003822. DOI: 10.1016/j.bspc.2022.103863.
[15]
HOSSAIN M B, SMHS IQBAL, ISLAM M M, et al. Transfer learning with fine-tuned deep CNN ResNet50 model for classifying COVID-19 from chest X-ray images[J/OL]. Info Med Unlo, 2022, 30: 100916 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S235291482200065X. DOI: 10.1016/j.imu.2022.100916.
[16]
AAMIR M, RAHMAN Z, DAYO Z A, et al. A deep learning approach for brain tumor classification using MRI images[J/OL]. Comput Electr Eng, 2022, 101: 108105 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S0045790622003603. DOI: 10.1016/j.compeleceng.2022.108105.
[17]
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/OL]. Neuroscience Info, 2021, 1: 100013 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S2772528621000133. DOI: 10.1016/j.neuri.2021.100013.
[18]
KHAIRANDISH M O, SHARMA M, JAIN V, et al. A hybrid CNN-SVM threshold segmentation approach for tumor detection and classification of MRI brain images[J]. IRBM, 2021, 43(4): 290-299. DOI: 10.1016/j.irbm.2021.06.003.
[19]
DEVI R S, PERUMAL B, RAJASEKARAN M P. A hybrid deep learning based brain tumor classification and segmentation by stationary wavelet packet transform and adaptive kernel fuzzy c means clustering[J/OL]. Adv Eng Softw, 2022, 170: 103146 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S0965997822000576. DOI: 10.1016/j.advengsoft.2022.103146.
[20]
NEELIMA G, CHIGURUKOTA D R, MARAM B, et al. Optimal DeepMRSeg based tumor segmentation with GAN for brain tumor classification[J/OL]. Biomed Signal Proces, 2022, 74: 103537 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S1746809422000593. DOI: 10.1016/j.bspc.2022.103537.
[21]
BADŽA M M, BARJAKTAROVIĆ M Č. Classification of brain tumors from MRI images using a convolutional neural network[J/OL]. Appl Sci, 2020, 10(6): 1999. https://www.mdpi.com/2076-3417/10/6/1999. DOI: 10.3390/app10061999.
[22]
AURNA N F, YOUSUF M A, TAHER K A, et al. A classification of MRI brain tumor based on two stage feature level ensemble of deep CNN models[J/OL]. Comput Biol Med, 2022, 146: 105539 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S0010482522003316. DOI: 10.1016/j.compbiomed.2022.105539.
[23]
KOKKALLA S, KAKARLA J, VENKATESWARLU I B, et al. Three-class brain tumor classification using deep dense inception residual network[J]. Soft Comput, 2021, 25(13): 8721-8729. DOI: 10.1007/s00500-021-05748-8.
[24]
GHASSEMI N, SHOEIBI A, ROUHANI M. Deep neural network with generative adversarial networks pre-training for brain tumor classification based on MR images[J/OL]. Biomed Signal Proces, 2020, 57: 101678 [2022-08-10]. https://www.sciencedirect.com/science/article/pii/S1746809419302599. DOI: 10.1016/j.bspc.2019.101678.
[25]
KHAN AR, KHAN S, HAROUNI M, et al. Brain tumor segmentation using K‐means clustering and deep learning with synthetic data augmentation for classification[J]. Microsc Res Techniq, 2021, 84(7): 1389-1399. DOI: 10.1002/jemt.23694.

PREV Feasibility study of fast arterial spin labeling imaging in brain on 3.0 T MRI scanner
NEXT The value of SEMAC-VAT imaging in the post-operative imaging of spine reconstruction surgery with 3D-printed vertebral body
  


LINK


Tel & Fax: +8610-67113815    E-mail: editor@cjmri.cn