Diagnostic abilities of three CAD methods for assessing microcalcifications in mammograms and an aspect of equivocal cases decisions by radiologists

  • W. T. Hung
  • H. T. Nguyen
  • W. B. Lee
  • M. T. Rickard
  • B. S. Thornton
  • A. Blinowska
Scientific Papers


Radiologists use an “Overall impression” rating to assess a suspicious region on a mammogram. The value ranges from 1 to 5. They will definitely send a patient for biopsy if the rating is 4 or 5. They will send the patient for core biopsy when a rating of 3 (indeterminate) is given. We have developed three methods to aid diagnosis of cases with microcalcifications. The first two methods, namely, Bayesian and multiple logistic regression (with a special “cutting score” technique), utilise six parameter ratings which minimise subjectivity in characterising the microcalcifications. The third method uses three parameters (age of patient, uniformity of size of microcalcification and their distribution) in a multiple stepwise regression. For both training set and test set, all three methods are as good as the two radiologists in terms of percentages of correct classification. Therefore, all three proposed methods potentially can be used as second readers.

Key words

computer aided diagnosis bayesian method multiple logistic regression multiple stepwise regression microcalcifications 


  1. 1.
    Lanyi, M.,Diagnosis and Differential Diagnosis of Breast Calcifications, Springer-Verlag, 5,13, 1988.Google Scholar
  2. 2.
    Glasziou, P. P.,Mammographic screening in Australia, Medical Journal Australia, 167: 516–517, 1997.Google Scholar
  3. 3.
    Kocur, C. M., Rogers, S. K., Myers, L. R., Burns, T., Kabrisky, M., Hoffmeister, J. F., Bauer, K. W., and Steppe, S. M,Using neural networks to select wavelet features for breast cancer diagnosis, IEEE Engineering in Medicine and Biology, 95-104, 1996.Google Scholar
  4. 4.
    Jiang, Y.,A computer-aided diagnostic scheme for classification of clustered microcalcifications in mammograms, Medical Physics, 26(6): 1018, 1999 (Ph.D. Thesis Abstract).CrossRefGoogle Scholar
  5. 5.
    Schmidt, F., Sorantin, E., Szepesvari, C., Graif, E., Becker, M., Mayer, H. and Hartwagner, K.,An automatic method for the identification and interpretation of clustered microcalcifications in mammograms, Physics in Medicine and Biology, 44: 1231–1243, 1999.CrossRefPubMedGoogle Scholar
  6. 6.
    Veldkamp, W. J. H., Karssemeijer, N., Otten, J. D. M. and Hendriks, J. H. C. L.,Automated classification of clustered microcalcifications into malignant and benign types, Medical Physics, 27(11): 2600–2608, 2000.CrossRefPubMedGoogle Scholar
  7. 7.
    Tourassi, G. D., Markey, M. K., Lo, J. Y. and Floyd Jr, C.E.,A neural network approach to breast cancer diagnosis as a constraint satisfaction problem, Medical Physics, 28(5): 804–811, 2001.CrossRefPubMedGoogle Scholar
  8. 8.
    Schmidt, R. A. and Nishikawa, R. M.,Clinical use of digital mammography: the present and the prospects, Journal of Digital Imaging, 8(1), SUPPL 1: 74–79, 1995.CrossRefPubMedGoogle Scholar
  9. 9.
    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 Sanjay-Gopal, S.,Improvement of radiologists’ characterization of mammographic masses by using computer-aided diagnosis: An ROC Study, Radiology, 212: 817–827, 1999.PubMedGoogle Scholar
  10. 10.
    Shen, L., Rangayyan, R. M., and Leo Desautels, J. E., Application of shape analysis to mammographic microcalcifications, IEEE Transactions. on Medical Imaging, 13: 263–74, 1994.CrossRefPubMedGoogle Scholar
  11. 11.
    Nguyen, H.,Neural Networks for Classifying Microcalcifications in Mammograms, Consulting Report to Foundation for Australian Resources, 1997.Google Scholar
  12. 12.
    Blinowska, A., Chatellier, G., Wojtasik, A. and Bernier, J.,Diagnostica — A Bayesian decision-aid system — applied to hypertension diagnosis, IEEE Transactions on Biomedical Engineering, 40(3): 230–235, 1993.CrossRefPubMedGoogle Scholar
  13. 13.
    Bland, M.,An Introduction to Medical Statistics, Oxford University Press, 1987.Google Scholar
  14. 14.
    Berger, J. O.,Statistical Decision Theory and Bayesian Analysis, Springer-Verlag, 1985.Google Scholar
  15. 15.
    Hosmer, D. W. and Lemeshow, S.,Applied Logistic Regression, New York, Wiley Press, 1989.Google Scholar
  16. 16.
    Beck, J. R. and Shultz, E. K.,The use of relative operating characteristics (ROC) curves in test performance evaluation, Archives of Pathology and Laboratory Medicine, 110: 13–20, 1986.PubMedGoogle Scholar
  17. 17.
    MedCal — statistical software for biomedical research version 7.1 (http://www.medcal.be/).Google Scholar

Copyright information

© Australasian College of Physical Scientists and Engineers in Medicine 2003

Authors and Affiliations

  • W. T. Hung
    • 1
  • H. T. Nguyen
    • 1
  • W. B. Lee
    • 2
  • M. T. Rickard
    • 3
  • B. S. Thornton
    • 1
  • A. Blinowska
    • 4
  1. 1.Faculty of EngineeringKey University Research Centre for Health TechnologiesSydneyAustralia
  2. 2.BreastScreen NSW-WesternSydneyAustralia
  3. 3.BreastScreen NSW-Central & EasternSydneyAustralia
  4. 4.INSERMHospital BroussaiParisFrance

Personalised recommendations