부산대학교 도서관

A serial study on 32 patients with bone metastases following cancer of the breast or prostate was performed over three years. Up to ten sets of images (average of four) per patient were obtained during this period using 99Tcmmethylene diphosphonate as the radiopharmaceutical. Ninety-three paired serial images of individual lesions were qualitatively assessed for change by three physicians in nuclear medicine and the results were compared with the quantitative results from computer analysis. The reproducibility of the quantitative approach was determined by the analysis of 20 paired lesions by three physicists. It was found that quantitative changes in uptake of less than 20% between images were generally not detected by the medical observers; a change of 41% had only a 95% probability of being identified as change by the physicians. Although much more reproducible in determining changes in individual lesions, the quantitative approach was found to be inferior to the qualitative assessment of overall change in the majority of cases which involve multiple lesions.The basic assumption that uptake varies proportionally with progression of the bone lesion is discussed and is considered in some instances to be untenable. The conclusion is drawn that the determination of progression from changes of uptake in longstanding lesions is uncertain and is subsidiary in importance to the detection of new lesions.