Region-Based Prediction and Quality Measurements for Medical Image Compression

This paper presents a prediction-based compression algorithm for medical image containing region of interest. The medical image on the whole consumes a lot of memory space which makes them difficult for storage and transmission. In a medical image with only a particular part needed for diagnosis, the important decision that has to be made is whether to go for block compression or region-based compression. Here region-based compression plays a vital role since a particular region alone can be preserved and the other regions can be compressed in a lossy way. Such methods are of great interest in tele-radiology applications with large storage requirements. Here, the quality of compression can be measured by capturing the size of the selected regions. A new method for calculating total compression ratio and total bits per pixel is proposed for such selective image compression algorithms. Since selected area of medical images is compressed lossless, the performance of the proposed system is compared with other lossless compression algorithms. The results showed comparatively good performance.