EVALUATING QUALITY OF COMPRESSED MEDICAL IMAGES - SNR, SUBJECTIVE RATING, AND DIAGNOSTIC-ACCURACY

Compressing a digital image can facilitate its transmission, storage, and processing. As radiology departments become increasingly digital, the quantities of their imaging data are forcing consideration of comp ression in picture archiving and communication systems (PACS) and evol ving teleradiology systems. Significant compression is achievable only by lossy algorithms, which do not permit the exact recovery of the or iginal image. This loss of information renders compression and other i mage processing algorithms controversial because of the potential loss of quality and consequent problems regarding liability, but the techn ology must be considered because the alternative is delay, damage, and loss in the communication and recall of the images. How does one deci de if an image is good enough for a specific application, such as diag nosis, recall, archival, or educational use? We describe three approac hes to the measurement of medical image quality: signal-to-noise ratio (SNR), subjective rating, and diagnostic accuracy. We compare and con trast these measures in a particular application, consider in some dep th recently developed methods for determining diagnostic accuracy of l ossy compressed medical images, and examine how good the easily obtain able distortion measures like SNR are at predicting the more expensive subjective and diagnostic ratings. The examples are of medical images compressed using predictive pruned tree-structured vector quantizatio n, but the methods can be used for any digital image processing that p roduces images different from the original for evaluation.