Lossless Medical Image Compression Based on Roi

Abstract

In today’s scenario, there is a continual increase in size of datasets of medical imaging
modality. As a result, there is a stupendous demand for huge medical imaging records, which
in turn requires more data space and memory. This implies for the need of suitable
compression scheme with high compression ratio. Usually, SPHIT and JPEG are applied to
medical images. Both the algorithms have high compression ratio, limitations been relating
to quality depreciation. Therefore a compromise has to be reached focusing mainly on
maintaining the image quality, information and also storage size. An algorithm is proposed
taking into fold both the benefits of lossy & lossless compression techniques. The input is any
type of medical image. This medical image is divided into region of interest (ROI), non
region of interest (NROI) and background. The proposed algorithm treats these ROI, NROI
and Background with various degree of interest, giving highest precision (lossless) to ROI as
compared to ROI and background.