Advertisement

Genetic Algorithms for Feature Selection in Computer-Aided Diagnosis

  • B. Sahiner
  • H. P. Chan
  • N. Petrick
Part of the Studies in Fuzziness and Soft Computing book series (STUDFUZZ, volume 96)

Abstract

One of the important practical applications of computer vision techniques is computer-aided diagnosis (CAD) in medical imaging. It has been shown that CAD can improve the accuracy of breast cancer detection and characterization by radiologists on mammograms. In this chapter, we discuss an important step — feature selection — in classifier design for CAD algorithms. Feature selection reduces the dimensionality of an available feature space and is therefore often used to prevent over-parameterization of a classifier. Many feature selection techniques have been proposed in the literature. We will illustrate the usefulness of genetic algorithms (GAs) for feature selection by comparing GA with a commonly used sequential selection method. A brief introduction to the GA is given and several examples using GA feature selection for the characterization of mammographic lesions are discussed. The examples illustrate the design of a fitness function for optimizing classification accuracy in terms of the receiver operating characteristics of the classifier, the dependence of GA performance on its evolution parameters, and the design of a fitness function tailored to a specific classification task.

Keywords

Genetic Algorithm Feature Selection Feature Space Receiver Operating Characteristic Curve Feature Selection Method 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    Parkin, D.M., Pisani, P., and Ferlay, J. (1999), “Global cancer statistics,” CA Cancer J. Clin., vol. 49, pp. 33–64.PubMedCrossRefGoogle Scholar
  2. [2]
    Greenlee, R.T., Murray, T., Bolden, S., and Wingo, P.A. (2000), “Cancer statistics, 2000,” CA Cancer J. Clin., vol. 50, pp. 7–33.PubMedCrossRefGoogle Scholar
  3. [3]
    Feig, S.A. and Hendrick, R.E. (1993), Risk, Benefit, and Controversies in Mammographic Screening, in Syllabus: A categorical Course in Physics Technical Aspects of Breast Imaging, A.G. Haus and M.J. Yaffe, Editors. Radiological Society of North America, Inc, Oak Brook, IL. pp. 119–135.Google Scholar
  4. [4]
    Duffy, S.W. and Tabar, L. (2000), “Screening mammograpgy reevaluated,” Lancet, vol. 355, pp. 747–748.PubMedCrossRefGoogle Scholar
  5. [5]
    Smart, C.R., Hendrick, R.E., Rutledge, J.H., and Smith, R.A. (1995), “Benefit of mammography screening in women ages 40 to 49 years: current evidence from randomized controlled trials,” Cancer, vol. 75, pp. 1619–1626.PubMedCrossRefGoogle Scholar
  6. [6]
    Martin, J.E., Moskowitz, M., and Milbrath, J.R. (1979), “Breast cancer missed by mammography,” AJR, vol. 132, pp. 737–739.PubMedCrossRefGoogle Scholar
  7. [7]
    Wallis, M.G., Walsh, M.T., and Lee, J.R. (1991), “A review of false negative mammography in a symptomatic population,” Clin. Radiol., vol. 44, pp. 13–15.PubMedCrossRefGoogle Scholar
  8. [8]
    Bird, R.E., Wallace, T.W., and Yankaskas, B.C. (1992), “Analysis of cancers missed at screening mammography,” Radiol., vol. 184, pp. 613–617.Google Scholar
  9. [9]
    Harvey, J.A., Fajardo, L.L., and Innis, C.A. (1993), “Previous mammograms in patients with impalpable breast carcinomas: retrospective vs blinded interpretation,” AJR, vol. 161, pp. 1167–1172.PubMedCrossRefGoogle Scholar
  10. [10]
    Adler, D.D. and Helvie, M.A. (1992), “Mammographic biopsy recommendations,” Curr. Op. Radiol., vol. 4, pp. 123–129.Google Scholar
  11. [11]
    Kopans, D.B. (1991), “The positive predictive value of mammography,” AJR, vol. 158, pp. 521–526.CrossRefGoogle Scholar
  12. [12]
    Shtern, F., Stelling, C., Goldberg, B., and Hawkins, R. (1995), “Novel technologies in breast imaging: national Cancer Institute perspective,” Society of Breast Imaging Conference, pp. 153–156.Google Scholar
  13. [13]
    Chan, H.P., Doi, K., Vyborny, C.J., Schmidt, R.A., Metz, C.E., Lam, K.L., Ogura, T., Wu, Y., and MacMahon, H. (1990), “Improvement in radiologists’ detection of clustered microcalcifications on mammograms. The potential of computer-aided diagnosis,” Invest. Radiol., vol. 25, pp. 1102–1110.PubMedCrossRefGoogle Scholar
  14. [14]
    Kegelmeyer, W.P., Pruneda, J.M., Bourland, P.D., Hillis, A., Riggs, M.W., and Nipper, M.L. (1994), “Computer-aided mammographic screening for spiculated lesions,” Radiol., vol. 191, pp. 331–337.Google Scholar
  15. [15]
    Chan, H.-P., Sahiner, B., Helvie, M.A., Petrick, N., Roubidoux, M.A., Wilson, T.E., Adler, D.D., Paramagul, C., Newman, J.S., and Gopal, S.S. (1999), “Improvement of radiologists’ characterization of mammographic masses by computer-aided diagnosis: an ROC study,” Radiol., vol. 212, pp. 817–827.Google Scholar
  16. [16]
    Jiang, Y., Nishikawa, R.M., Schmidt, R.A., Metz, C.E., Giger, M.L., and Doi, K. (1999), “Improving breast cancer diagnosis with computer-aided diagnosis,” Acad. Rad., vol. 6, pp. 22–33.CrossRefGoogle Scholar
  17. [17]
    Kilday, J., Palmieri, F., and Fox, M.D. (1993), “Classifying mammographic lesions using computer-aided image analysis,” IEEE Trans. Med. Img., vol. 12, pp. 664–669.CrossRefGoogle Scholar
  18. [18]
    Chan, H.P., Wei, D., Helvie, M.A., Sahiner, B., Adler, D.D., Goodsitt, M.M., and Petrick, N. (1995), “Computer-aided classification of mammographic masses and normal tissue: linear discriminant analysis in texture feature space,” Phys. Med. Biol., vol. 40, pp. 857–876.PubMedCrossRefGoogle Scholar
  19. [19]
    McNitt-Gray, M.F., Huang, H.K., and Sayre, J.W. (1995), “Feature selection in the pattern classification problem of digital chest radiograph segmentation,” IEEE Trans. Med. Img., vol. 14, pp. 537–547.CrossRefGoogle Scholar
  20. [20]
    Sahiner, B., Chan, H.P., Petrick, N., Wagner, R.F., and Hadjiiski, L. (2000), “Feature selection and classifier performance in computer-aided diagnosis: the effect of finite sample size.,” Med. Phys., vol. 27, pp. 1509–1522.PubMedCrossRefGoogle Scholar
  21. [21]
    Jain, A. and Zongker, D. (1997), “Feature selection: Evaluation, application, and small sample size performance,” IEEE Trans. Pat. Anal. Mach. Intell., vol. 19, pp. 153–158.CrossRefGoogle Scholar
  22. [22]
    Ferri, F.J., Pudil, P., Hatef, M., and Kittler, J. (1994), “Comparative study of techniques for large-scale feature selection,” Pattern Recognition in Practice, vol. IV, pp. 403–413.Google Scholar
  23. [23]
    Box, G.E.P. (1957), “Evolutionary operation: a method for increasing industrial productivity,” Appl. Stat., vol. 6, pp. 81–101.CrossRefGoogle Scholar
  24. [24]
    Holland, J.H. (1962), “Outline for a logical theory of adaptive systems,” J. Assoc. Comput. Mach., vol. 3, pp. 297–314.CrossRefGoogle Scholar
  25. [25]
    Fogel, L.J., Owens, A.J., and Walsh, M.J. (1966), Artificial Intelligence Through Simulated Evolution, Wiley, New York.Google Scholar
  26. [26]
    Forrest, S. (1993), “Genetic algorithms: principles of natural selection applied to computation,” Science, vol. 261, pp. 872–878.PubMedCrossRefGoogle Scholar
  27. [27]
    Jain, A.K., Duin, R.P.W., and Mao, J. (2000), “Statistical pattern recognition: a review,” IEEE Trans. Pat. Anal. Mach. Intell., vol. 22, pp. 4–37.CrossRefGoogle Scholar
  28. [28]
    Cover, T.M. and Campenhpout, J.M.V. (1977), “On the possible orderings in the measurement selection problem,” IEEE Trans. Sys. Man. and Cybern., vol. 7, pp. 657–661.CrossRefGoogle Scholar
  29. [29]
    Cover, T.M. (1974), “The best two independent measurements are not the two best,” IEEE Trans. Sys. Man. and Cybern., vol. 6, pp. 116–117.CrossRefGoogle Scholar
  30. [30]
    Wu, Y., Giger, M.L., Doi, K., Vyborny, C.J., Schmidt, R.A., and Metz, C.E. (1993), “Artificial neural networks in mammography: application to decision making in the diagnosis of breast cancer,” Radiol., vol. 187, pp. 81–87.Google Scholar
  31. [31]
    Meisel, W.S. (1972), Computer-oriented approaches to pattern recognition, Academic Press, New York.Google Scholar
  32. [32]
    Duda, R.O. and Hart, P.E. (1973), Pattern Classification and Scene Analysis, Wiley, New York.Google Scholar
  33. [33]
    Raudys, S.J. and Jain, A.K. (1991), “Small sample size effects in statistical pattern recognition: recommendations for practitioners,” IEEE Trans. Pat. Anal. Mach. Intell., vol. 13, pp. 252–264.CrossRefGoogle Scholar
  34. [34]
    Narendra, P.M. and Fukunaga, K. (1977), “A branch and bound algorithm for feature subset selection,” IEEE Trans. Comput., vol. 26, pp. 917–922.CrossRefGoogle Scholar
  35. [35]
    Siedlecki, W. and Sklansky, J. (1989), “A note on genetic algorithm for large-scale feature selection,” Patt. Recog. Let., vol. 10, pp. 335–347.CrossRefGoogle Scholar
  36. [36]
    ]Brill, F., Brown, D., and Martin, W. (1992), “Fast genetic selection of features for neural network classifiers,” IEEE Trans. Neural Net., vol. 3, pp. 324–328.CrossRefGoogle Scholar
  37. [37]
    ]Kuncheva, L.I. and Jain, L.C. (1999), “Nearest neighbor classifier: simulataneous editing and feature selection,” Patt. Recog. Let., vol. 20, pp. 1149–1156.CrossRefGoogle Scholar
  38. [38]
    Kudo, M. and Sklansky, J. (2000), “Comparison of algorithms that select features for pattern classifiers,” Patt. Recog., vol. 33, pp. 2541.CrossRefGoogle Scholar
  39. [39]
    Kudo, M. and Sklansky, J. (1998), “A comparative evaluation of medium- and large-scale feature selectors for pattern classifiers,” Kybernetika, vol. 34, pp. 429–434.Google Scholar
  40. [40]
    Sahiner, B., Chan, H.P., Petrick, N., Wei, D., Helvie, M.A., Adler, D.D., and Goodsitt, M.M. (1996), “Image feature selection by a genetic algorithm: Application to classification of mass and normal breast tissue on mammograms,” Med. Phys., vol. 23, pp. 1671–1684.PubMedCrossRefGoogle Scholar
  41. [41]
    Sahiner, B., Chan, H., Petrick, N., Helvie, M., and Goodsitt, M. (1998), “Design of a high-sensitivity classifier based on a genetic algorithm: application to computer-aided diagnosis,” Phys. Med. Biol., vol. 43, pp. 2853–2871.PubMedCrossRefGoogle Scholar
  42. [42]
    Chan, H.P., Sahiner, B., Lam, K.L., Petrick, N., Helvie, M.A., Goodsitt, M.M., and Adler, D.D. (1998), “Computerized analysis of mammographie microcalcifications in morphological and texture feature space,” Med. Phys., vol. 25, pp. 2007–2019.PubMedCrossRefGoogle Scholar
  43. [43]
    Zheng, B., Chang, Y.-H., Wang, X.-H., Good, W.F., and Gur, D. (1999), “Feature selection for computerized mass detection in digitized mammograms by using a genetic algorithm,” Acad. Rad., vol. 6, pp. 327–332.CrossRefGoogle Scholar
  44. [44]
    Yamany, S.M., Khiani, K.J., and Faraq, A.A. (1997), “Application of neural networks and genetic algorithms in the classification of endothelial cells,” Patt. Recog. Let., vol. 18, pp. 1205–1210.CrossRefGoogle Scholar
  45. [45]
    Handels, H., Ross, T., Kreusch, J., Wolff, H.H., and Poppl, S.J. (1999), “Feature selection for optimized skin tumor recognition using genetic algorithms,” Art. Intel. Med., vol. 16, pp. 283–297.CrossRefGoogle Scholar
  46. [46]
    Kupinski, M.A. and Anastasio, M.A. (1999), “Multiobjective genetic optimization of diagnostic classifiers with implications for generating receiver operating characteristic curves,” IEEE Trans. Med. Img., vol. 18, pp. 675–685.CrossRefGoogle Scholar
  47. [47]
    Anastasio, M.A., Yoshida, H., Nagel, R., Nishikawa, R.M., and Doi, K. (1998), “A genetic algorithm-based method for optimizing the performance of a computer-aided diagnosis scheme for detection of clustered microcalcifications in mammograms,” Med. Phys., vol. 25, pp. 1613–1620.PubMedCrossRefGoogle Scholar
  48. [48]
    Gudmundsson, M., El-Kwae, E.A., and Kabuka, M.R. (1998), “Edge detection in medical images using a genetic algorithm,” IEEE Trans. Med. Img., vol. 17, pp. 469–474.CrossRefGoogle Scholar
  49. [49]
    Fujita, H., Hara, T., Jing, X., Matsumoto, T., Yoshimura, H., and Seki, K. (1995), “Automated detection of lung nodules by using a genetic algorithm technique in chest radiography,” Radiol., vol. 197 (P), pp. 426–426.Google Scholar
  50. [50]
    Pena-Reyes, C.A. and Sipper, M. (2000), “Evolutionary computation in medicine: an overview,” Art. Intel!. Med., vol. 19, pp. 1–23.CrossRefGoogle Scholar
  51. [51]
    Lachenbruch, P.A. (1975), Discriminant Analysis, Hafner Press, New York.Google Scholar
  52. [52]
    Metz, C.E. (1986), “ROC methodology in radiologie imaging,” Invest. Radiol., vol. 21, pp. 720–733.PubMedCrossRefGoogle Scholar
  53. [53]
    Dorfman, D. and Alf Jr, E. (1969), “Maximum likelihood estimation of parameters of signal detection theory and determination of confidence intervals-rating method data.,” J. Math. Psych., vol. 6, pp. 487–496.CrossRefGoogle Scholar
  54. [54]
    Metz, C.E., Herman, B.A., and Shen, J.H. (1998), “Maximum-likelihood estimation of receiver operating characteristic (ROC) curves from continuously-distributed data,” Stat. Med.,vol. 17, pp. 1033–1053.Google Scholar
  55. [55]
    Tatsuoka, M.M. (1988), Multivariate Analysis, Techniques for Educational and Psychological Research, 2nd ed. Macmillan, New York.Google Scholar
  56. [56]
    Norusis, M.J. (1993), SPSS for Windows Release 6 Professional Statistics, SPSS Inc., Chicago, IL.Google Scholar
  57. [57]
    Wei, D., Chan, H.P., Helvie, M.A., Sahiner, B., Petrick, N., Adler, D.D., and Goodsitt, M.M. (1995), “Classification of mass and normal breast tissue on digital mammograpms: multiresolution texture analysis,” Med. Phys., vol. 22, pp. 1501–1513.PubMedCrossRefGoogle Scholar
  58. [58]
    Chan, H.P., Wei, D., Lam, K.L., Sahiner, B., Helvie, M.A., Adler, D.D., and Goodsitt, M.M. (1995), “Classification of malignant and benign microcalcifications by texture analysis,” Med. Phys., vol. 22, pp. 938.Google Scholar
  59. [59]
    Chan, H.P., Sahiner, B., Wei, D., Helvie, M.A., Adler, D.D., and Lam, K.L. (1995), “Computer-aided diagnosis in mammography: Effect of feature classifier on characterization of microcalcifications,” Radiol., vol. 197 (P), pp. 425.Google Scholar
  60. [60]
    Chan, H.P., Sahiner, B., Petrick, N., Helvie, M.A., Leung, K.L., Adler, D.D., and Goodsitt, M.M. (1997), “Computerized classification of malignant and benign microcalcifications on mammograms: texture analysis using an artificial neural network,” Phys. Med. Biol., vol. 42, pp. 549–567.PubMedCrossRefGoogle Scholar
  61. [61]
    Chan, H.P., Niklason, L.T., Ikeda, D.M., and Adler, D.D. (1992), “Computer-aided diagnosis in mammography: detection and characterization of microcalcifications,” Med. Phys., vol. 19, pp. 831.Google Scholar
  62. [62]
    Chan, H.P., Wei, D., Lam, K.L., Lo, S.-C.B., Sahiner, B., Helvie, M.A., and Adler, D.D. (1995), “Computerized detection and classification of microcalcifications on mammograms,” Proc. SPIE Med. Img., vol. 2434, pp. 612–620.CrossRefGoogle Scholar
  63. [63]
    Sahiner, B., Chan, H.P., Petrick, N., Wei, D., Helvie, M.A., Adler, D.D., and Goodsitt, M.M. (1996), “Classification of mass and normal breast tissue: a convolution neural network classifier with spatial domain and texture images,” IEEE Trans. Med. Img., vol. 15, pp. 598–610.CrossRefGoogle Scholar
  64. [64]
    Haralick, R.M., Shanmugam, K., and Dinstein, I. (1973), “Texture features for image classification,” IEEE Trans. Sys. Man. and Cybern., vol. SMC-3, pp. 610–621.Google Scholar
  65. [64]
    ]Metz, C.E., Wang, P.L., and Kronman, H.B. (1984), “A new approach for testing the significance for differences between ROC curves measured from correlated data,” in: Deconinck, F. (Ed.), Information Processing in Medical Imaging, The Hague, Martinus Nijhoff, pp. 432–445.CrossRefGoogle Scholar
  66. [66]
    Vyborny, C.J. and Giger, M.L. (1994), “Computer vision and artificial intelligence in mammography,” AJR, vol. 162, pp. 699–708.PubMedCrossRefGoogle Scholar
  67. [67]
    Petrick, N., Chan, H.P., Wei, D., Sahiner, B., Helvie, M.A., and Adler, D.D. (1996), “Automated detection of breast masses on mammograms using adaptive contrast enhancement and texture classification,” Med. Phys., vol. 23, pp. 1685–1696.PubMedCrossRefGoogle Scholar
  68. [68]
    Wei, D., Chan, H.P., Petrick, N., Sahiner, B., Helvie, M.A., Adler, D.D., and Goodsitt, M.M. (1997), “False-positive reduction technique for detection of masses on digital mammograms: global and local multiresolution texture analysis,” Med. Phys., vol. 24, pp. 903–914.PubMedCrossRefGoogle Scholar
  69. [69]
    ]Rumelhart, D.E., Hinton, G.E., and Williams, R.J. (1986), “Learning internal representation by error propagation,” in: Rumelhart, D.E. and McClelland, J.L. (Eds.), Parallel Distributed Processing, vol. 1, MIT Press, Cambridge, MA.Google Scholar
  70. [70]
    Hermann, G., Janus, C., Schwartz, I.S., Krivisky, B., Bier, S., and Rabinowitz, J.G. (1987), “Nonpalpable breast lesions: Accuracy of prebiopsy mammographie diagnosis,” Radiol., vol. 165, pp. 323–326.Google Scholar
  71. [71]
    Hall, F.M., Storella, J.M., Silverstond, D.Z., and Wyshak, G. (1988), “Nonpalpable breast lesions: recommendations for biopsy based on suspicion of carcinoma at mammography,” Radiol., vol. 167, pp. 353.Google Scholar
  72. [72]
    Jacobson, H.G. and Edeiken, J. (1990), “Biopsy of occult breast lesions: analysis of 1261 abnormalities,” JAMA, vol. 263, pp. 2341–2343.CrossRefGoogle Scholar
  73. [73]
    ]Jiang, Y., Metz, C.E., and Nishikawa, R.M. (1996), “A receiver operating characteristic partial area index for highly sensitive diagnostic tests,” Radiol., vol. 201, pp. 745–750.Google Scholar
  74. [74]
    Sahiner, B., Chan, H.P., Petrick, N., Helvie, M.A., and Goodsitt, M.M. (1998), “Computerized characterization of masses on mammograms: the rubber band straightening transform and texture analysis,” Med. Phys., vol. 25, pp. 516–526.PubMedCrossRefGoogle Scholar
  75. [75]
    Weszka, J.S., Dyer, C.R., and Rosenfeld, A. (1976), “A comparative study of texture measures for terrain classification,” IEEE Trans. Sys. Man. and Cybern., vol. 6, pp. 269–285.CrossRefGoogle Scholar
  76. [76]
    Galloway, M.M. (1975), “Texture classification using gray level run lengths,” Comp. Graph. Img Proc., vol. 4, pp. 172–179.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • B. Sahiner
  • H. P. Chan
  • N. Petrick

There are no affiliations available

Personalised recommendations