Evaluation of irreversible compression of digitized posterior-anterior chest radiographs

The purpose of this article is to assess lossy image compression of digitized chest radiographs using radiologist assessment of anatomic structures and numerical measurements of image accuracy. Forty posterior-anterior (PA) chest radiographs were digitized and compressed using an irreversible wavelet technique at 10, 20, 40, and 80:1. These were presented in a blinded fashion with an uncompressed image for A-B comparison of 11 anatomic structures as well as overall quality assessments. Mean error, root-mean square (RMS) error, maximum pixel error, and number of pixels within 1% of original value were also computed for compression ratios from ratios from 5:1 to 80:1. We found that at low compression (10:1) there was a slight preference for compressed images. There was no significant difference at 20:1 and 40:1. There was a slight preference on some structures for the original compared with 80:1 compressed images. Numerical measures showed high image faithfulness, both in terms of number of pixels that were within 1% of their original value, and by the average error for all pixels. Our findings suggest that lossy compression at 40:1 or more can be used without perceptible loss in the representation of anatomic structures. On this finding, we will do a receiver-operator characteristic (ROC) analysis of nodule detection in lossy compressed images using 40:1 compression.