THE USE OF CONTEXTUAL INFORMATION IN THE REVERSIBLE COMPRESSION OF MEDICAL IMAGES

Methods for the reversible compression of (gray scale) medical images consist of two consecutive steps, viz., decorrelation and coding %1]. The coding of decorrelated image pixels in these methods is customarily accomplished by means of Huffman coding along with an implicit memoryless model. In this paper, we investigate the use of conditioning events (or contexts) in improving the performances of known compression methods by building a source model with multiple contexts to code the decorrelated pixels. Three methods for reversible compression, viz., DPCM (differential pulse code modulation), WHT (Walsh-Hadamard transform), and HINT (hierarchical interpolation), employing, respectively, predictive decorrelation, transform decorrelation, and multiresolution decorrelation are considered. It is shown that the performances of these methods can be enhanced significantly, sometimes even up to 40%, by using contexts. The enhanced DPCM method is found to perform the best for MR and UT (ultrasound) medical images; the enhanced WHT method is found to be the best for X-Ray images. The source models used in the enhanced methods employ several hundred contexts; the initial set of contexts is selected based on the gradients along two orthogonal directions in the original (or transformed) image; additional contexts are then adaptively selected by means of a context splitting procedure based on the estimated value of a pixel. The probability estimates of the decorrelated pixel values under each context are obtained using the so called cumulative adaptive technique. Coding is accomplished by means of arithmetic coding.