ABSTRACT
This chapter focuses on a novel algorithm for medical image compression, especially for magnetic resonance imaging (MRI) images. In this chapter, an efficient image compression technique constructed by integrating the proposed histogram indexed dictionary (HID) into the standard Huffman encoding has been proposed for MRI of brain images. The discrete cosine transform is a technique that converts a signal into elementary frequency components. A sliding neighborhood filtering is an operation that is performed on one coefficient at a time, with the value of any given coefficient in the output matrix being determined by the application of an algorithm to the values of the corresponding input coefficient's neighborhood. Zig-zag scanning is not a property of DCT itself, but it is a part of the transform-based coding. Quantization reduces the number of bits per samples. The peak signal-to-noise ratio (PSNR) measures the difference in pixel value between two images, and is widely used to measure the quality of compressed or reconstructed images.
