Medical Image Compression Based on Set Partitioning in Hierarchical Trees Using Quantized Coefficients of Self Organizing Feature Map for Mr Images

Medical imaging plays a vital role in medical diagnosis. These medical images available in hospitals and medical organizations occupy a lot of space. The massive use of digitized images has led to the compression allowing economical storage and fast data transfer. Over the years, JPEG compression schemes based on Discrete Cosine Transform have been proposed and standardized. The input image has to be blocked which results in blocking artifacts. In recent years wavelet transform has gained widespread acceptance in image compression. Many compression algorithms using wavelets like EZW, SPIHT and SPECK have been proposed and they can be used for lossy or lossless compression. SPIHT is an efficient compression algorithm which has better performance over the others. In this paper, SPIHT based medical image compression algorithm using the quantized coefficients of Self Organizing Feature Map (SOFM) was brought in and proved to have better performance over existing methods.