Advertisement

Medical and Biological Engineering and Computing

, Volume 19, Issue 5, pp 597–604 | Cite as

Detectability index, likelihood ratio and diagnostic output as psychophysical measures for evaluation of medical image processing

  • J. P. J. de Valk
  • E. G. J. Eijkman
  • J. Vintcent
Article

Abstract

Signal-detection theory was used to study methods of measuring changes in diagnostic quality from medical image processing. Detectability index and likelihood ratio are applicable for purposes of comparison but they do not give an indication of the confidence with which a diagnosis has been made. Therefore a new measure for evaluation, the diagnostic output, is defined which incorporates this important aspect. This study describes experiments where four observers judge pictures of heart scintigrams in different stages of processing involving variations of resolution and brightness. The emphasis is on methods for evaluation of the diagnostic value of picture processing. Our conclusion is that the influence of the studied image-processing techniques can be well described using signaldetection theory, while the diagnostic output seems to be a well suited measure to represent diagnostic properties of stimulus and observer.

Key words

Diagnostic image processing Diagnostic output Signal-detection theory 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Buddemeyer, E. U., Bachrach, S. L. andMitchell, T. G. (1979) Instrumentation in nuclear cardiology. InCardiovascular nuclear medicine, C. V.Mosby Co., St. Louis, 3–45.Google Scholar
  2. Eijkman, E. G. J. (1980) Psychophysics. InHandbook of psychonomics I, North-Holland Publ. Co., Amsterdam, 303–363.Google Scholar
  3. Fu, K. S. (1976) Processing of chest X-ray images by computer. InDecision making and medical care North-Holland Publ. Co., Amsterdam, 271–286.Google Scholar
  4. Fu, K. S. (1980) Picture syntax. InLecture notes in computer science. Pictorial information systems, Springer Verlag, Berlin, 104–127.Google Scholar
  5. Eden, G. andGelsema, S. (1979) Investigation of multidimensional data using the interactive pattern analysis system ISPAHAN.Pattern Recognition 11, 391–399.CrossRefGoogle Scholar
  6. Gallagher, J. H., Preston, D. F., Robinson, R. G., Herrin, W. F., Servoss, W. andFritz, S. (1977) Image processing and pattern recognition in nuclear medicine.IEEE Trans. on Inf. Processing and Pattern Rec.,55, 55–60.Google Scholar
  7. Gelsema, E. S. andEden, G. (1980) Mapping algorithms in ISPAHAN.Pattern Recognition,12, 127–136.CrossRefGoogle Scholar
  8. Goodenough, D. J. (1976) Assessment of image quality of diagnostic imaging system. InMedical images: formation perception and measurement,Hay,G. A. (Ed.), Wiley, New York, 263–277.Google Scholar
  9. Green, D. M. andSwets, J. A. (1966)Signal detection and psychophysics, Wiley, New York.Google Scholar
  10. Grey, D. R. andMorgan, B. J. T. (1972) Some aspects of ROC-curve-fitting: normal and logistic models.J. Math. Psychol.,9, 128–139.CrossRefGoogle Scholar
  11. McKelvie, S. J. (1978) Graphic rating scales—How many categories?Br. J. Psychol.,69, 185–202.Google Scholar
  12. Metz, C. E., Goodenough, D. J. andRossmann, K. (1973) Evaluation of receiver operating characteristic curve data in terms of information theory, with applications in radiography.Radiology,109, 297–303.Google Scholar
  13. Metz, C. E., Starr, S. J. andLusted, L. B. (1976) Quantitative evaluation of visual detection performance in medicine: ROC-analysis and determination of diagnostic benefit. InMedical images: formation, perception and measurement.Hay, G. A., (Ed.), Wiley, New York, 220–241.Google Scholar
  14. Nadler, M. (1976) Effective and cost-effective real time picture operators for medical imagery. InDecision making and medical care, North-Holland Publ. Co, Amsterdam, 259–279.Google Scholar
  15. Netravali, A. N. andLimb J. O. (1980) Picture coding: a review.Proc. IEEE,68, 366–406.CrossRefGoogle Scholar
  16. Sharp, P. F. (1976) Psychophysical assessment of the penformance of clinical radio-isotope imaging devices. InMedical images: formation perception and measurement.Hay, G. A. (Ed.), Wiley, New York., 242–260.Google Scholar
  17. Sharp, P. F., Chesser, R. B. andSmith, D. B. (1980) Assessment of scintigram quality.IEEE Proc. Part E,127, 185–192.Google Scholar
  18. Simpson, A. J. andFitter, M. J. (1973) What is the best index of detectability?Psychological Bulletin,80, 481–488.CrossRefGoogle Scholar
  19. Starr, S. J., Metz, C. E. Lusted, L. B. andGoodenough, D. J. (1975) Visual detection and localisation of radiographic images.Radiology 116, 533–538.Google Scholar
  20. Todd-Pokpopek, A. E. (1976) Image processing in nuclear medicine: an examination of the quest for a measure of clinical quality. InMedical images: formation, perception and measurement.Hay, G. A. (Ed.), Wiley, New York, 278–292.Google Scholar

Copyright information

© IFMBE 1981

Authors and Affiliations

  • J. P. J. de Valk
    • 1
  • E. G. J. Eijkman
    • 1
  • J. Vintcent
    • 1
  1. 1.Laboratory of Medical Physics and BiophysicsUniversity of NijmegenNijmegenThe Netherlands

Personalised recommendations