INTRODUCTION: An ROC analysis was carried out in order to determine the reliability of digital luminescence radiography review at a PC and this was compared with a radiological work station and with X-ray film on a viewing box. MATERIAL AND METHOD: 54 chest images obtained by digital luminescence radiography were selected, 31 of these contained small pulmonary nodules. In order to evaluate critical detail, five images of a phantom showing round foci were used. Five radiologists examined these, using a Siemens Magic View work station, a PC with proprietary software (ViewMed) and X-ray films on a viewing box. Image processing of the work station used the standard clinical application. ViewMed performs linear scaling of grey levels to 8 Bit. The results were examined statistically by means of a t-test. RESULTS: As far as the chest images were concerned there was no significant difference in the diagnostic value of these methods. There was, however, a highly significant loss of diagnostic information with respect to the round focus phantom when using the PC compared with the other methods. CONCLUSION: In the configuration in which it was used, the PC should not be relied on as a primary means of examination since critical details cannot always be seen. In routine use these play a subordinate role and there was no significant diagnostic loss where the chest images were concerned. We expect that by improvements in the frequency and contrast processing the PC accuracy will be considerably increased.
«
INTRODUCTION: An ROC analysis was carried out in order to determine the reliability of digital luminescence radiography review at a PC and this was compared with a radiological work station and with X-ray film on a viewing box. MATERIAL AND METHOD: 54 chest images obtained by digital luminescence radiography were selected, 31 of these contained small pulmonary nodules. In order to evaluate critical detail, five images of a phantom showing round foci were used. Five radiologists examined these, u...
»