Digital Radiography Enhancement by Nonlinear Multiscale Processing

Today's digital radiography systems mostly use unsharp maskinglike enhancement algorithms based on splitting input images into two or three frequency channels. This method allows fine detail enhancement as well as processing of global contrast (harmonization). However, structures of medium size are not accessible. In extension of a standard algorithm of such type, we develop and test a new enhancement algorithm based on hierarchically repeated unsharp masking, resulting in a multiscale architecture allowing consistent access to structures of all sizes. Our algorithm decomposes a radiograph by a pyramid-architecture, dividing it into eight or more channels representing structures of different sizes, known as "scales." At each scale, weakly contrasting structures are then enhanced by suitable nonlinear processing. We emphasize two points: first, backward compatibility to the standard algorithm which is used routinely in clinical practice. This allows reuse of current parametrization know-how as well as a smooth transition from current to new processing. Second, our enhancement is noise-resistant in the sense that it prevents unacceptable noise amplification. A prototype implementation of the algorithm is undergoing trials in the clinical routine of radiology departments of major German hospitals. Results strongly indicate the superior performance and high acceptance of the new processing.