Visually lossless adaptive compression of medical images

A method for encoding medical images to visually lossless quality is proposed. Visually lossless coding refers to the compression of images without evoking any perceptible degradation in image quality through the suppression of visible distortions. The proposed coding structure is based on the set partition in hierarchical trees (SPIHT) coder (A. Said et al., June 1996). Added to this is an adaptive vision model designed to identify and removal visually irrelevant information in images prior to encoding. Simulation results of the proposed coder have shown compression improvements over the LOCO (M. Weinberger et al., 1996) coder without any deterioration in image fidelity.