AI helps you reading Science
AI generates interpretation videos
AI extracts and analyses the key points of the paper to generate videos automatically
AI parses the academic lineage of this thesis
AI extracts a summary of this paper
All experimental cases were randomly divided into five groups, and one group was chosen as a testing group by turns while the remaining four groups were used to train the support vector machine with Genetic algorithms
Combining support vector machine with genetic algorithm to classify ultrasound breast tumor images.
Computerized Medical Imaging and Graphics, no. 8 (2012): 627-633
To promote the classification accuracy and decrease the time of extracting features and finding (near) optimal classification model of an ultrasound breast tumor image computer-aided diagnosis system, we propose an approach which simultaneously combines feature selection and parameter setting in this study.
PPT (Upload PPT)
- Breast cancer has been the main cause of death for women globally in recent years. According to statistics, in 2009, 192,370 women in the United States were expected to be diagnosed with breast cancer, and 40,170 deaths were attributed to the disease .
- Detection of breast cancer usually consists of a physical examination, imaging, and biopsy .
- Biopsy is the best way to accurately determine whether the tumor is benign or malignant.
- It is invasive and is much more expensive than other detection methodologies.
- Breast cancer has been the main cause of death for women globally in recent years
- All experimental cases were randomly divided into five groups, and one group was chosen as a testing group by turns while the remaining four groups were used to train the support vector machine with Genetic algorithms
- We evaluated the ultrasound breast tumor images with the new six-dimensional feature vector
- We considered evaluating breast tumor images using all 30 features
- The feature selection step is implemented with the genetic algorithm and identifies the significant features used to evaluate the breast tumor images
- Parameters setting for a support vector machine are often designed quite differently due to the unique characteristics of the data
- A total of 210 ultrasound images that had been pathologically proven were used to evaluate the CAD system.
- The ultrasound image database included 120 benign breast tumor images and 90 malignant ones.
- The patients’ ages ranged from 18 to 64 years old and only one image from each patient is contained in the database.
- This study was approved by the local ethics committee and informed consent was waived.
- The patient was in supine position with arms extended overhead.
- No acoustic standoff pad was used with any of the cases
- This study used the 5-fold cross-validation proposed by Salzberg . That is, all experimental cases were randomly divided into five groups, and one group was chosen as a testing group by turns while the remaining four groups were used to train the SVM with GA.
The proposed approach was implemented on a PC with an AMD Athlon 64 1.8 GHz processor, Windows XP operating system, and the Visual C++ 6.0 development environment.
- The libSVM  is used in the proposed approach.
- For SVM with GA, the authors set the following parameter values: population size is 20, crossover rate is 90%, mutation rate is 10%, the maximum number of solutions evaluated is 10,000 for GA, while the range of parameter C is 1–30,000 and the range of parameter is 0.001–1 for the SVM.
- Each component f[i] of the feature vector was defined as follows:
- The authors proposed an automatic diagnostic system that uses practical texture and morphological features to effectively distinguish between benign and malignant lesions of the breast in this study.
- The authors considered evaluating breast tumor images using all 30 features.
- The feature selection step is implemented with the genetic algorithm and identifies the significant features used to evaluate the breast tumor images.
- Trial-and-error method seems to be the most common way to identify the nearoptimal parameters of an SVM.
- It is time-consuming and does not guarantee the better result.
- The authors use the GA to find the near-optimal parameters C and of the SVM
- Table1: Features selected among five runs
- Table2: Objectively indices result of the different approaches/features
- Table3: Computational time required for various approaches to obtain their best models (time of feature extraction is not included)
- Table4: Paired t-tests on average classification accuracy for various approaches
- Table5: Time comparison of calculating features of all breast tumor images with different approaches
- This research was partially supported by the National Science Council of the Republic of China (Taiwan) under Contract no
- Jemal A, Murray T, Samuels A, Ghafoor A, Ward E, Thun MJ. Cancer statistics. CA Cancer J Clin 2009;59:225–49.
- Esserman LJ. New approaches to the imaging, diagnosis, and biopsy of breast lesions. Cancer J 2002;8:S1–14.
- Chang RF, Wu WJ, Moon WK, Chen DR. Automatic ultrasound segmentation and morphology based diagnosis of solid breast tumors. Breast Cancer Res Treat 2005;89:179–85.
- Chen DR, Chang RF, Huang YL. Computer-aided diagnosis applied to US of solid breast nodules by using neural networks. Radiology 1999;213:407–12.
- Chen ST, Hsiao YH, Huang YL, Kuo SJ, Tseng HS, Wu HK, et al. Comparative analysis of logistic regression, support vector machine and artificial neural network for the differential diagnosis of benign and malignant solid breast tumors by the use of three-dimensional power Doppler imaging. Korean J Radiol 2009;10:464–71.
- Chen WM, Chang RF, Moon WK, Chen DR. Breast cancer diagnosis using three-dimensional ultrasound and pixel relation analysis. Ultrasound Med Biol 2003;29:1027–35.
- Chen CY, Chiou HJ, Chou SY, Chiou SY, Wang HK, Chou YH. Computer-aided diagnosis of soft-tissue tumors using sonographic morphologic and texture features. Acad Radiol 2009;16:1531–8.
- Chen DR, Huang YL, Lin SH. Computer-aided diagnosis with textural features for breast lesions in sonograms. Comput Med Imaging Graph 2011;35:220–6.
- Boyd NF, Guo H, Martin LJ. Mammographic density and the risk and detection of breast cancer. N Engl J Med 2007;356:227–36.
- Saslow D, Boetes C, Burke W. American Cancer Society guidelines for breast screening with MRI as an adjunct to mammography. CA Cancer J Clin 2007;57:75–89.
- Griebsh I, Brown J, Boggis C. Cost-effectiveness of screening with contrast enhanced magnetic resonance imaging vs X-ray mammography of women at a high familial risk of breast cancer. Br J Cancer 2006;95:801–10.
- Leach MO, Boggis CR, Dixon AK. Screening with magnetic imaging and mammography of a UK population at high familial risk of breast cancer: a prospective multicentre cohort study (MARIBS). Lancet 2006;365:1769–78.
- Stavros AT, Thickman D, Rapp CL, Dennis MA, Parker SH, Sisney GA. Solid breast nodules: use of sonography to distinguish between benign and malignant lesions. Radiology 1995;196:123–34.
- Chang RF, Wu WJ, Moon WK, Chou YH, Chen DR. Support vector machines for diagnosis of breast tumors on US images. Acad Radiol 2003;10:189–97.
- Lin SW, Chen SC. PSOLDA: a particle swarm optimization approach for enhancing classification accuracy rate of linear discriminant analysis. Appl Soft Comput 2009;9:1008–15.
- Holland JH. Genetic algorithms. Sci Am 1992;267:44–50.
- Goldberg DE. Genetic algorithms in search, optimization, and machine learning. Boston, MA: Addison-Wesley Longmam Publishing Co., Inc.; 1989.
- Bhanu B, Lin Y. Genetic algorithm based feature selection for target detection in SAR images. Image Vis Comput 2003;21:591–608.
- Hong JH, Cho SB. Efficient huge-scale feature selection with speciated genetic algorithm. Pattern Recogn Lett 2006;27:143–50.
- Lu J, Zhao T, Zhang Y. Feature selection based-on genetic algorithm for image annotation. Knowl Based Syst 2008;21:887–91.
- Chen GY, Xie WF. Pattern recognition with SVM and dual-tree complex wavelets. Image Vis Comput 2007;25:960–6.
- Chang RF, Wu WJ, Moon WK, Chen DR. Improvement in breast tumor discrimination by support vector machines and speckle-emphasis texture analysis. Ultrasound Med Biol 2003;29:679–86.
- Wang XY, Yang HY, Cui CY. An SVM-based robust digital image watermarking against desynchronization attacks. Signal Process 2008;88:2193–205.
- Wong WT, Hsu SH. Application of SVM and ANN for image retrieval. Eur J Oper Res 2006;173:938–50.
- Vapnik VN. The nature of statistical learning theory. New York: SpringerVerlag; 1995.
- Liu Z, Zhang D, Yan JQ, Li QL, Tang QL. Classification of hyperspectral medical tongue images for tongue diagnosis. Comput Med Imaging Graph 2007;31:672–8.
- Rezatofighi SH, Soltanian-Zadeh H. Automatic recognition of five types of white blood cells in peripheral blood. Comput Med Imaging Graph 2011;35:333–43.
- Deguchi K, Izumitani T, Hontani H. Detection and enhancement of line structures in an image by anisotropic diffusion. Pattern Recogn Lett 2002;23:1399–405.
- Czerwinski RN, Jones DL, O’Brien WD. Detection of lines and boundaries in speckle images – application to medical ultrasound. IEEE Trans Med Imaging 1999;18:126–36.
- Blayvas I, Bruckstein A, Kimmel R. Efficient computation of adaptive threshold surfaces for image binarization. Pattern Recogn 2006;39:89–101.
- Zhou H, Yuan Y, Lin F, Liu T. Level set image segmentation with Bayesian analysis. Neurocomputing 2008;71:1994–2000.
- Christianini N, Shawe-Taylor J. An introduction to support vector machines and other kernel-based learning methods. Cambridge: Cambridge University Press; 2000.
- Zollner FG, Emblem KE, Schad LR. Support vector machines in DSC-based glioma imaging: suggestions for optimal characterization. Magn Reson Med 2010;64(4):1230–6.
- Liao Y, Fang SC, Nuttle HLW. A neural network model with bounded-weights for pattern classification. Comput Oper Res 2004;31:1411–26.
- Lin HT, Lin CJ. A study on sigmoid kernels for SVM and the training of non-PSD kernels by SMO-type methods. Technical report, University of National Taiwan, Department of Computer Science and Information Engineering, vol. 3; 2003. pp. 1–32.
- Müller KR, Mike S, Rätsch G, Tsuda K, Schölkopf B. An introduction to kernelbased learning algorithms. IEEE Trans Neural Netw 2001;12:181–201.
- Pardo M, Sberveglieri G. Classification of electronic nose data with support vector machines. Sens Actuators B: Chem 2005;107:730–7.
- Gen M, Cheng R. Genetic algorithms and engineering design. NY: John Wiley & Sons Co., Inc.; 1997.
- Salzberg SL. On comparing classifiers: pitfalls to avoid and a recommended approach. Data Mining Knowl Discov 1997;1:317–27.
- Hsu C-W, Chang C-C, Lin C-J, A practical guide to support vector classification. Technical report, University of National Taiwan, Department of Computer Science and Information Engineering; July 2003. http://www.csie.ntu.edu.tw/∼cjlin/libsvm/.
- Marcano-Cedeno A, Quintanilla-Dominguez J, Andina D. WBCD breast cancer database classification applying artificial metaplasticity neural network. Expert Syst Appl 2011;38:9573–9. Wen-Jie Wu is an assistant professor in Department of Information Management, Chang Gung University, Taiwan. He received the B.S. degree in computer science and engineering from Tatung University, Taipei, Taiwan, in 1996, and the M.S. degree and Ph.D. degree in computer science and information engineering from Nation Chung Cheng University, Chiayi, Taiwan, in 1998 and 2003, respectively. His research interests include image processing, medical computer-aided diagnosis system, and data mining.
- Shih-Wei Lin is an associate professor in Department of Information Management, Chang Gung University, Taiwan. He received his Ph.D. in industrial management from the National Taiwan University of Science and Technology in 2002. His current research interests include scheduling and data mining. His papers have appeared in Computers and Operations Research, European Journal of Operational Research, Journal of the Operational Research Society, International Journal of Production Research, International Journal of Advanced Manufacturing Technology, Knowledge and Information Systems, Applied Soft Computing, Applied Intelligence, and Expert Systems with Applications, etc.
- Woo Kyung Moon is a professor in Department of Radiology, College of Medicine, Seoul National University, Korea. He received the M.D. degree from Seoul National University College of Medicine, Seoul, Korea, in 1989, and the Ph.D. degree in radiology science from Seoul National University, Seoul, Korea, in 1999. His research interests include breast imaging and intervention, computer-aided diagnosis, and molecular imaging.