Interpretation accuracy of a CCD film digitizer

J Digit Imaging. 2002;15 Suppl 1(2):57-63. doi: 10.1007/s10278-002-5099-5. Epub 2002 Mar 21.


There are no significant differences between the interpretations of radiographic images resulting from digitizing films using a recently developed CCD unit and the readings of the original films as measured by accuracy, sensitivity, specificity and ROC analysis. Digital imaging is rapidly becoming the basis of radiology practice resulting in the gradual elimination of conventional film examinations. Related to this trend is the ability of radiologists to provide reliable interpretations of imaging examinations using film digitization and soft copy display. An acceptable system must provide high-quality digital images to achieve the levels of diagnostic accuracy that are comparable to the interpretation of film. This study is an effort to contribute to the acceptance of digital imaging by testing the hypothesis.- The authors selected 120 radiographic examinations from the departmental film files which included chest, abdominal, extremity and other common procedures. Half of the cases contained specified abnormalities and half did not have such findings. All of the examinations had a high degree of diagnostic difficulty. The films were digitized on the CCD equipment and each of the four board certified radiologists who participated in the study interpreted each of the 120 cases, half on original film and half in the digitized format on a high resolution workstation. No radiologist read the same examination more than once. Data collection included variables such as perceived image quality, diagnostic difficulty and interpretation confidence. Accuracy measures were calculated and an ROC analysis was performed. As of this date, the preliminary results indicate acceptance of the hypothesis that there is no significant difference between the accuracy of film readings and the interpretations of the digitized images on a soft copy display. Differences noted in perceived image quality among readers were also not significant. The final results, including an ROC analysis, will be available in March 2002 when a paper is submitted for publication in the Journal of Digital Imaging. During the transition from conventional film practice to digital radiology there is a need to convert previous films to a digital format for comparison with new digital examinations and to facilitate expert reading and consultation on film procedures performed at distant sites. Diagnostic quality and cost effectiveness are key factors in the acceptability of film digitizers to meet these requirements. The preliminary results of this study are encouraging in that another option may be available to radiologists and administrators responsible for choosing equipment for the changing pratice of medical imaging.

MeSH terms

  • Data Display*
  • Humans
  • Observer Variation
  • ROC Curve
  • Radiographic Image Enhancement / instrumentation*
  • Radiographic Image Enhancement / standards
  • Sensitivity and Specificity
  • X-Ray Intensifying Screens