Implementation and Performance Assessment of Gradient Edge Detection Predictor for Reversible Compression of Biomedical Images

Abstract

Technological advancement of medical imaging techniques are progressing constantly, dealing with images of increasing resolutions. In hospitals, medical imaging techniques like X-rays, magnetic resonance imaging and computed tomography etc. are of high resolution consuming large storage space. Such high resolution of medical images transmitted over the network utilizes large bandwidth that often results in degradation of image quality. So, compression of images is only a solution for efficient archival and communication of medical images. Predictive based coding technique is explored in this paper for medical image compression as it performs well for lossless compression. This paper presents a comparative investigation on 2D predictor’s coding efficiency and complexity on CT images. It was observed that among 2D predictors Gradient Edge Detection (GED) predictor gave better results than Median Edge Detector (MED) and DPCM. GED predictor at proper threshold value achieved approximately same results in terms of various performance metrics as Gradient Adaptive Predictor (GAP) though it is less complex.