Clinical Image Evaluation of Mammograms: A National Survey

ABSTRACT

PURPOSE: The goal of this study was to survey the overall quality of mammographic images in Korea. MATERIALS AND METHODS: A total of 598 mammographic images collected from 257 hospitals nationwide were reviewed in terms of eight image quality categories, namely positioning, compression, contrast, exposure, sharpness, noise, artifacts, and examination identification, and rated on a five-point scale: (1=severe deficiency, 2=major deficiency, 3=minor deficiency, 4=good, 5=best). Failure was defined as the occurrence of more than four major deficiencies or one severe deficiency (score of 1 or 2). The results were compared among hospitals of varying kinds, and common problems in clinical image quality were identified. RESULTS: Two hundred and seventeen mammographic images (36.3%) failed the evaluation. Poor images were found in descending order of frequency, at The Society for Medical Examination (33/69, 47.8%), non-radiology clinics (42/88, 47.7%), general hospitals (92/216, 42.6%), radiology clinics (39/102, 38.2%), and university hospitals (11/123, 8.9%) (p<0.01, Chi-square test). Among the 598 images, serious problems which occurred were related to positioning in 23.7% of instances (n=142) (p<0.01, Chi-square test), examination identification in 5.7% (n=34), exposure in 5.4% (n=32), contrast in 4.2% (n=25), sharpness in 2.7% (n=16), compression in 2.5% (n=15), artifacts in 2.5% (n=15), and noise in 0.3% (n=2). CONCLUSION: This study showed that in Korea, 36.3% of the mammograms examined in this sampling had important image-related defects that might have led to serious errors in patient management. The failure rate was significantly higher in non-radiology clinics and at The Society for Medical Examination than at university hospitals.