Journal of Biomedical Engineering Research (대한의용생체공학회:의공학회지)

The Korean Society of Medical and Biological Engineering (대한의용생체공학회)
권호 표지
DOI QR코드

DOI QR Code

Malignant and Benign Classification of Liver Tumor in CT according to Data pre-processing and Deep running model

CT영상에서의 AlexNet과 VggNet을 이용한 간암 병변 분류 연구

  • Choi, Bo Hye (Bio-Medical Engineering, College of Health Science, Gachon University) ;
  • Kim, Young Jae (Department of Biomedical Engineering, College of Medicine, Gachon University) ;
  • Choi, Seung Jun (Department of Radiology, Gachon University Gil Hospital) ;
  • Kim, Kwang Gi (Department of Biomedical Engineering, College of Medicine, Gachon University)
  • 최보혜 (가천대학교 보건과학대학 의용생체공학과) ;
  • 김영재 (가천대학교 의과대학 의공학교실) ;
  • 최승준 (가천대학교 길병원 영상의학과) ;
  • 김광기 (가천대학교 의과대학 의공학교실)
  • Received : 2018.06.18
  • Accepted : 2018.10.30
  • Published : 2018.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

Liver cancer is one of the highest incidents in the world, and the mortality rate is the second most common disease after lung cancer. The purpose of this study is to evaluate the diagnostic ability of deep learning in the classification of malignant and benign tumors in CT images of patients with liver tumors. We also tried to identify the best data processing methods and deep learning models for classifying malignant and benign tumors in the liver. In this study, CT data were collected from 92 patients (benign liver tumors: 44, malignant liver tumors: 48) at the Gil Medical Center. The CT data of each patient were used for cross-sectional images of 3,024 liver tumors. In AlexNet and VggNet, the average of the overall accuracy at each image size was calculated: the average of the overall accuracy of the $200{\times}200$ image size is 69.58% (AlexNet), 69.4% (VggNet), $150{\times}150$ image size is 71.54%, 67%, $100{\times}100$ image size is 68.79%, 66.2%. In conclusion, the overall accuracy of each does not exceed 80%, so it does not have a high level of accuracy. In addition, the average accuracy in benign was 90.3% and the accuracy in malignant was 46.2%, which is a significant difference between benign and malignant. Also, the time it takes for AlexNet to learn is about 1.6 times faster than VggNet but statistically no different (p > 0.05). Since both models are less than 90% of the overall accuracy, more research and development are needed, such as learning the liver tumor data using a new model, or the process of pre-processing the data images in other methods. In the future, it will be useful to use specialists for image reading using deep learning.

Keywords

OOSCB@_2018_v39n6_229_f0001.png 이미지

그림 1. 영상 전처리 과정의 예. Fig. 1. Image data pre-processing process : (a) Original Image(12bit) (b) WW/WL(8bit) (c) ROI (d) Resize(150X150).

OOSCB@_2018_v39n6_229_f0002.png 이미지

그림 2. AlexNet : 5개 컨볼루션 층, 3개의 완전 연결된 층. Fig. 2. AlexNet : 5 convolutional layer, 3 fully-connected layer.

OOSCB@_2018_v39n6_229_f0003.png 이미지

그림 3. Vgg Net 구조. Fig. 3. Vgg Net Architecture.

OOSCB@_2018_v39n6_229_f0004.png 이미지

그림 4. AlexNet을 이용한 학습시간 결과. Fig. 4. Learning time(s) using Alexnet.

OOSCB@_2018_v39n6_229_f0005.png 이미지

그림 5. VggNet을 이용한 학습시간 결과. Fig 5. Learning time(s) using VggNet.

표 1. 이미지 크기, 배치 크기, 반복 횟수의 차이에 따른 AlexNet과 VggNet 결과 비교 Table 1. Comparison of AlexNet and VggNet results due to difference in image size, batch size and epoch

OOSCB@_2018_v39n6_229_t0001.png 이미지

표 2. AlexNet과 VggNet의 학습 시간 비교 결과 Table 2. AlexNet and VggNet results through analysis of learning time

OOSCB@_2018_v39n6_229_t0002.png 이미지

References

  1. J. Shim, "Current Status of Research on Liver Cancer in Korea," The Korean Liver Cancer Journal, vol. 17, no. 2, pp.105-112, 2017. https://doi.org/10.17998/jlc.17.2.105
  2. https://www.cancer.go.kr/lay1/S1T639C640/contents.do, National Cancer Information Center, accessed on June. 13, 2018
  3. M.A. Lee, "Special Review: Screening and surveillance of hepatocellular carcinoma," Korean Journal of Medicine, vol. 79, no. 3, pp.224-230, 2010.
  4. T. S.Furey, N. Cristianini, N. Duffy, D. W.Bendnarski, M. Schummer and D. Haussler, "Support vector machine classification and validation of cancer tissue samples using microarray expression data," Bioinformatics, vol. 16, no. 10, pp. 906-914, 2000. https://doi.org/10.1093/bioinformatics/16.10.906
  5. M. Langkvist, A. Kiselev, M. Alirezaie and A. Loutfi, "Classification and Segmentation of Satellite Orthoimagery Using Convolutional Neural Networks," Remote Sensing, vol. 8, no. 4, pp. 329, 2016. https://doi.org/10.3390/rs8040329
  6. J. Zhang, K.K. Ma, M.H. Er and V. Chong, "Tumor Segmentation from Magnetic Resonance Imaging by Learning via one-class support vector machine," in International Workshop on Adavanced Image Technology (IWAIT'04), Singapore, Jan 2004, pp. 207-211.
  7. S. Zhang, L. Yang, C. Peng and M. Wu, "Logistic regression analysis of risk factors for postoperative recurrence of spinal tumors and analysis of prognostic factors," Oncology Letters, vol. 15, no. 2, pp. 1716-1722, 2018.
  8. H.P. Chan, B. Sahiner, N. Petrick, M.A. Helvie, K.L. Lam, D.D. Adler and M.M. Goodsitt, "Computerized classification of malignant and benign microcalcifications on mammograms: texture analysis using an artificial neural network," Physics in Medicine & Biology, vol. 42, no. 3, pp. 549, 1997. https://doi.org/10.1088/0031-9155/42/3/008
  9. V. Gulshan, L. Peng, M. Coram, M. C.Stumpe, D. Wu, A. Narayanaswamy, S. Venugopalan, K. Widner, T. Madams, J.Cuadros, R. Kim, R. Raman, P. C.Nelson, J. L.Mega and D. R.Webster, "Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs," JAMA, vol. 316, no. 22, pp. 2402-2410, 2016. https://doi.org/10.1001/jama.2016.17216
  10. H.C. Park, E.J. Lee, Y.J. Kim, S.W. Lee, "Colonoscopy Image Classification Using Convolutional Neural Networks," The Korean Institute of Next Generation Computing, vol. 14, no. 1, pp. 89-98, 2018.
  11. H.W. Kim, B.N. Kim, J.E. Lee, W.S. Jang and S.K. Yoo, "Performance Improvement of Convolutional Neural Network for Pulmonary Nodule Detection," Journal of Biomedical Engineering Research, vol. 38, no. 5, pp. 237-241, 2017. https://doi.org/10.9718/JBER.2017.38.5.237
  12. J. H. Lee, H. S. Ahn, D. H. Choi, and K. S. Tae, "Evaluation on the Usefulness of X-ray Computer-Aided Detection (CAD) System for Pulmonary Tuberculosis (PTB) using SegNet", Journal of Biomedical Engineering Research, vol. 38, no.1, pp. 25-31, 2017 https://doi.org/10.9718/JBER.2017.38.1.25
  13. L. Gattinoni, P. Caironi, P. Pelosi and L. R.Goodman, "What Has Computed Tomography Taught Us about the," AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE, vol. 164, no. 9, pp. 1701-1711, 2001. https://doi.org/10.1164/ajrccm.164.9.2103121
  14. S. Sabouri., A. Khatami, P. Azadeh, J. Ghoroubi and Gh. Azimi, "Adding Liver Window Setting to the Standard Abdominal CT Scan Protocol: Is It Useful?," Iranian Journal of Radiology, vol. 5, no. 2, pp. 65-70, 2008.
  15. A. Krizhevsky, I. Sutskever and G. E.Hinton, "ImageNet Classification with Deep Convolutional Neural Networks," in In Advances in neural information processing systems, 2012, pp. 1097-1105.
  16. K. Simonyan and A. Zisserman, "Very deep convolutional networks for large-scale image recognition," in arXiv preprint arXiv:1409.1556, Apr. 2014.