The legal-clinical conflict of image compression
An interesting session on image compression was held at ECR Thursday March 4:
“Image compression: Acceptable? Unacceptable? Any decision yet?”.
Dr David Koff presented results from the impressively thorough national Canadian evaluation*. Seen as a summary of the Radiology community’s view on image compression, the session reflected a refreshingly pragmatic view on how to use image compression for the benefit of clinical effectiveness. There are, however, still problematic differences between clinical and legal aspects of image compression.
The session showed that a fundamental insight now seems to be undisputed: that the data explosion is faster than the price reduction of storage space and bandwidth. Thus, compression and other advanced data management methods continue to be crucial.
The update on medico-legal evaluations of image compression is definitely food for thought. A positive sign is the common legal interpretation that compression is allowed as long as the diagnostic assessment is not affected. Clinical effectiveness is in focus as it should be. On the other hand, the legal conclusion is also that the archived image must be identical to the one viewed, which means that any compression used must be applied before the radiologist’s review. Here two interests are in conflict.
The interpretation does make sense since applying heavy compression after review may destroy evidence of malpractice. But from a clinical standpoint the goal should always be to have all data gathered accessible to the radiologist’s review. The most problematic “data loss” scenario of this type is probably not compression, but when a thin slice stack is scanned but only reconstructed thick slices are sent to PACS.
There is, actually, an approach that may resolve the legal-clinical conflict: Always provide all image data to the radiologist and let him/her view the data in whatever way that ensures a solid diagnostic assessment. Let the professional judgment decide which pixels to view, at which compression ratio and in which resolution. Then archive exactly what has been shown, no more, no less. This would result in a high compression ratio while ensuring that the provenance of the diagnostic work is retained for legal purposes.
* Pan-Canadian Evaluation of Irreversible Compression Ratios (“Lossy” Compression) for Development of National Guidelines, Journal of Digital Imaging, Vol 22, No 6 (December), 2009: pp 569-578
