CHAPTER 6

Image Compression

A compressed medical image can **reduce** the image size (see Chapter 2, table 2.1) as well as shorten the transmission time and decrease an image’s storage requirement. But a compressed image may compromise the image’s original quality and affect its diagnostic value. This chapter describes some compression techniques that are applicable to medical images, and their advantages and disadvantages. In addition recent trends on using image compression in clinical environment are presented.

6.1 TERMINOLOGY

The half-dozen definitions that follow are essential to an understanding of image compression/reconstruction.

*Original image in 2-D, 3-D, and 4-D.*A two-dimensional (2-D) original digital medical image*f*(*x*,*y*) is a nonnegative integer function where*x*and*y*can be from 0 to 255, 0 to 511, 0 to 1023, 0 to 2047, or 0 to 4095 (see Section 2.1). In three-dimensional (3-D) image,*f*(*x*,*y*,*z*) is a 3-D spatial data block, or*f*(*x*,*y*,*t*) with spatial 2-D and the time as the third dimension. In the case of four-dimensional image (4-D),*f*(*x*,*y*,*z*,*t*) is a sequence of multiple 3-D spatial data blocks collected at different times (see Sections 3.1, 4.1, and 5.1). Image compression of the original 2-D image (a rectangular array), 3-D image (a three-dimensional data block), or 4-D image data set (a time sequence of 3-D data blocks), respectively, is to compress it into a one-dimensional data file.*Transformed image.*The transformed image*F*(*u*,*v*) of the original 2-D image ...