Compression of Biomedical Images and Signals

Chapter 7

Compression of 2D Biomedical Images

7.1. Introduction

On a daily basis, large amounts of medical images are acquired using 2D acquisition imaging systems (e.g., vertebra and lung digital X rays, mammography). Moreover, it is possible to compress temporal sequences (i.e. 2D+t), volume sequences (i.e. 3D) or even spatio-temporal sequences (i.e. 3D+t) by encoding each image separately and independently of all others (i.e. in clinical routine, physicians do not always keep all images but instead select the most relevant and accurate ones). Thus, 2D compression is widely applied to medical images. It is also included in the DICOM format (described in Chapter 4), within various PACS.

This chapter is a review of some basic 2D compression methods which are frequently applied to medical images. Although the common compression techniques and the traditional standards of compression do apply to medical images, some specific methods applied to specific images have been specially developed in order to optimize both the compression rates and the quality of the re-constructed image.

This chapter is made up of three main parts. Section 7.2 will look at the compression of medical images using reversible methods (i.e. lossless). This will be followed by an examination of lossy techniques in section 7.3, and finally, progressive compression methods will be described in section 7.4. Thus, we will show that this type of compression is highly appropriate to the transmission of medical information. ...