跳到主要內容

臺灣博碩士論文加值系統

(18.97.14.82) 您好!臺灣時間:2025/02/15 01:19
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:Nguyen Ngoc Nhut
研究生(外文):NGUYEN NGOC NHUT
論文名稱:Classification of Brain Tumor MRI using Transfer Learning with EfficientNetv2
論文名稱(外文):Classification of Brain Tumor MRI using Transfer Learning with EfficientNetv2
指導教授:黃有評黃有評引用關係
指導教授(外文):HUANG, YO-PING
口試委員:姚立德李祖添蘇順豐吳炳飛
口試委員(外文):YAO, LEEHTERLEE, TSU-TIANSU, SHUN-FENGWU, BING-FEI
口試日期:2022-07-08
學位類別:碩士
校院名稱:國立臺北科技大學
系所名稱:電資學院外國學生專班(iEECS)
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2022
畢業學年度:110
語文別:英文
論文頁數:47
中文關鍵詞:deep learningbrain tumormagnetic resource imaging (MRI)transfer learningclassification
外文關鍵詞:deep learningbrain tumormagnetic resource imaging (MRI)transfer learningclassification
相關次數:
  • 被引用被引用:0
  • 點閱點閱:109
  • 評分評分:
  • 下載下載:18
  • 收藏至我的研究室書目清單書目收藏:0
Computer-aided diagnosis (CAD) is becoming increasingly important and can be applied to many medical fields. CAD can greatly support the radiologists in reducing the time and effort for the diagnosis. One of the medical fields which uses CAD is the computer-assisted classification of brain tumors that becomes easier when working with the MRI images. This is critical because brain tumors are among the leading causes of mortality worldwide, and the quick diagnosis of this pathology not only helps patients early know their disease but also helps doctors easily find the right treatment. This research focuses on classifying three categories of brain tumors: glioma, meningioma, and pituitary tumors. These types of brain tumors are worthy of investigation. Deep learning models have been recognized as important methods of modern innovation, or the development of modern technology. It helps people minimize the gap between human vision and the computer vision to increase the high performance in the work of humans in the field of image processing such as classification, segmentation, or detection. With the appearance of numerous networks for classification, particularly the EfficientNetv2, MobileNetv2, and ResNet50, are not only the networks with faster training and more network weights reduction, but also help obtain higher performance in the classification task. Transfer learning has become a machine learning technique that can help us reduce the amount of time needed to solve a problem or reduce the cost by reusing a previously trained network for a new task. Furthermore, the applications of transfer learning have gradually appeared in many research. Therefore, this research implemented a transfer learning method with EfficientNetv2 on two independent datasets to get the best performance, and we were able to achieve a result with over 97% accuracy. The performance was validated utilizing five-fold cross-validation, and the metrics included overall accuracy, precision, recall, and confusion matrices. The aim of this research is to develop a novel approach for classifying brain tumors to assist doctors in their diagnosis.
Computer-aided diagnosis (CAD) is becoming increasingly important and can be applied to many medical fields. CAD can greatly support the radiologists in reducing the time and effort for the diagnosis. One of the medical fields which uses CAD is the computer-assisted classification of brain tumors that becomes easier when working with the MRI images. This is critical because brain tumors are among the leading causes of mortality worldwide, and the quick diagnosis of this pathology not only helps patients early know their disease but also helps doctors easily find the right treatment. This research focuses on classifying three categories of brain tumors: glioma, meningioma, and pituitary tumors. These types of brain tumors are worthy of investigation. Deep learning models have been recognized as important methods of modern innovation, or the development of modern technology. It helps people minimize the gap between human vision and the computer vision to increase the high performance in the work of humans in the field of image processing such as classification, segmentation, or detection. With the appearance of numerous networks for classification, particularly the EfficientNetv2, MobileNetv2, and ResNet50, are not only the networks with faster training and more network weights reduction, but also help obtain higher performance in the classification task. Transfer learning has become a machine learning technique that can help us reduce the amount of time needed to solve a problem or reduce the cost by reusing a previously trained network for a new task. Furthermore, the applications of transfer learning have gradually appeared in many research. Therefore, this research implemented a transfer learning method with EfficientNetv2 on two independent datasets to get the best performance, and we were able to achieve a result with over 97% accuracy. The performance was validated utilizing five-fold cross-validation, and the metrics included overall accuracy, precision, recall, and confusion matrices. The aim of this research is to develop a novel approach for classifying brain tumors to assist doctors in their diagnosis.
Abstract i
Acknowledgments iii
Table of Contents iv
List of Tables vi
List of Figures vii
Chapter 1 Introduction 1
1.1 Background 1
1.2 Related Work 4
1.3 Research Objective 5
1.4 Thesis Development 5
Chapter 2 Literature Review 6
2.1 ResNet50 6
2.2 MobileNetv2 7
2.3 EfficientNetv2 8
2.4 Transfer Learning 12
2.5 K fold Cross Validation 13
Chapter 3 Proposed Methodology 15
3.1 Proposed Structure 15
3.2 Training Process 16
Chapter 4 Experiment Setup 18
4.1 Datasets 18
4.1.1 Figshare Dataset 18
4.1.2 Kaggle Dataset 19
4.2 Preprocessing 20
4.2.1 Normalization 20
4.2.2 Data Augmentation 21
4.3 Experiment Implementation 21
4.4 Hyperparameters 22
Chapter 5 Result and Discussion 24
5.1 Performance Metrics 24
5.2 K fold Cross Validation Result 25
5.3 Final Training Result 28
5.4 Final Evaluation Result 30
5.5 Discussion 42
5.6 Comparison with The State of Art 43
Chapter 6 Conclusions 45
References 46
[1]A. Rehman, S. Naz, M. I. Razzak, F. Akram, and M. Imran, “A deep learning-based framework for automatic brain tumors classification using transfer learning,” Circuits System Signal Processing, vol. 39, no. 2, pp.757-775, Feb. 2020.
[2]https://www.cancerresearchuk.org/about-cancer/brain-tumours (accessed Apr. 30, 2022).
[3]https://www.cancer.net/cancer-types/brain-tumor/statistics (accessed Apr. 30, 2022).
[4]N. Noreen, S. Palaniappan, A. Qayyum, I. Ahmad, M. Imran, and M. Shoaib, “A deep learning model based on concatenation approach for the diagnosis of brain tumor,” IEEE Access, vol. 8, no. 1, pp.55135-55144, Mar. 2020.
[5]P. Afshar, K. N. Plataniotis, and A. Mohammadi, “Capsule networks for brain tumor classification based on MRI images and course tumor boundaries,” in Proc. of Int. Conf. on Acoustics, Speech and Signal Processing (ICASSP), pp.1368-1372, Brighton, UK, May 2019.
[6]S. Deepak and P. M. Ameer, “Brain tumor classification using deep CNN features via transfer learning,” Computers in Biology and Medicine, vol. 111, no. 1, p. 103345, Aug. 2019.
[7]M. Sajjad, S. Khan, K. Muhammad, W. Wu, A. Ullah, and S. W. Baik, “Multi-grade brain tumor classification using deep CNN with extensive data augmentation,” Journal of Computational Science, vol. 30, no. 1, pp.174-182, Jan. 2019.
[8]Y. Gu, X. Lu, L. Yang, B. Zhang, D. Yu, Y. Zhao, and L. Gao et al., “Automatic lung nodule detection using a 3D deep convolutional neural network combined with a multi-scale prediction strategy in chest CTs,” Computers in Biology and Medicine, vol. 103, no. 1, pp.220-231, Dec. 2018.
[9]M. Yousefi, A. Krzyżak, and C.Y. Suen, “Mass detection in digital breast tomosynthesis data using convolutional neural networks and multiple instance learning,” Computers in Biology and Medicine, vol. 96, no. 1, pp.283-293, May 2018.
[10]L. Zhou, Z. Zhang, Y.C. Chen, Z.Y. Zhao, X.D. Yin, and H.B. Jiang, “A deep learning-based radiomics model for differentiating benign and malignant renal tumors,” Translational Oncology, vol. 12, no. 2, pp.292-300, Feb. 2019.
[11]C. Tan, F. Sun, T. Kong, W. Zhang, C. Yang, and C. Liu, “A survey on deep transfer learning,” in Proc. of Int. Conf. on Artificial Neural Networks, pp.270-279, Rhodes, Greece, Oct. 2018.
[12]M. Talo, U. B. Baloglu, Ö. Yıldırım, and U. R. Acharya, “Application of deep transfer learning for automated brain abnormality classification using MR images,” Cognitive Systems Research, vol. 54, no. 1, pp.176-188, May 2019.
[13]H. H. Sultan, N. M. Salem, and W. Al-Atabany, “Multi-classification of brain tumor images using deep neural network,” IEEE Access, vol. 7, no. 1, pp.69215-69225, May 2019.
[14]W. Ayadi, W. Elhamzi, I. Charfi, and M. Atri, “Deep CNN for brain tumor classification,” Neural Processing Letters, vol. 53, no. 1, pp.671-700, Feb. 2021.
[15]Z. N. K. Swati, Q. Zhao, M. Kabir, F.Ali, Z. Ali, S. Ahmed, and J. Lu, “Brain tumor classification for MR images using transfer learning and fine-tuning,” Computerized Medical Imaging and Graphics, vol. 75, no. 1 , pp.34-46, Jul. 2019.
[16]K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition, pp.770-778, Las Vegas, NV, USA, Jun. 2016.
[17]M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, and L.C. Chen, “Mobilenetv2: Inverted residuals and linear bottlenecks,” in Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), pp.4510-4520, Salt Lake City, UT, USA, Jun. 2018.
[18]M. Tan and Q. Le, “EfficientNet: Rethinking model scaling for convolutional neural networks,” in Proc. of the 36th Int. Conf. on Machine Learning, vol. 97, pp.6105-6114, Long Beach, NY, USA, Jun. 2019.
[19]J. Hu, L. Shen, and G. Sun, “Squeeze-and-excitation networks,” in Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), pp.7132-7141, Salt Lake City, UT, USA, Jun. 2018.
[20]M. Tan and Q. Le, “EfficientNetv2: Smaller models and faster training,” in Proc. of Int. Conf. on Machine Learning, vol. 139, pp.10096-10106, Virtual Event, Jul. 2021.
[21]https://ai.googleblog.com/2019/08/efficientnet-edgetpu-creating.html (accessed Apr. 30, 2022).
[22]S. Y. Lu, S. H. Wang, and Y. D. Zhang, “A classification method for brain MRI via MobileNet and feedforward network with random weights,” Pattern Recognition Letters, vol. 140, no. 1, pp.252-260, Dec. 2020.
[23]J. Han, M. Kamber, and J. Pei, Data Mining: Concepts and Techniques, third edition [M], The Morgan Kaufmann Series in Data Management Systems, vol. 5, no. 4, pp.83-124, Jun. 2011.
[24]J. Cheng, “Brain Tumor Dataset,” Figshare, Jun. 2017.
[25]J. Cheng, W. Huang, S. Cao, R. Yang, W. Yang, Z. Yun, and Z. Wang et al., “Enhanced performance of brain tumor classification via tumor region augmentation and partition,” PLOS ONE, vol. 10, no. 10, pp.1-13, May 2015.
[26]S. Bhuvaji, A. Kadam, P. Bhumkar, S. Dedge, and S. Kanchan, “Brain tumor classification (MRI),” Kaggle, May 2020.
[27]K. Maharana, S. Mondal, and B. Nemade, “A review: Data pre-processing and data augmentation techniques,” Global Transitions Proceedings, vol. 3, no. 1, pp.91-99, Jun. 2022.
[28]https://www.python.org/ (accessed Jan. 12, 2022).
[29]https://www.tensorflow.org/ (accessed Jan. 12, 2022).
[30]https://numpy.org/ (accessed Jan. 12, 2022).
[31]https://scikit-learn.org/stable/ (accessed Jan. 12, 2022).
[32]https://matplotlib.org/ (accessed Jan. 12, 2022).
[33]https://research.google.com/colaboratory/ (accessed Jan. 12, 2022).


QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關期刊