Compression of Biomedical Images and Signals

Chapter 3

Specificities of Physiological Signals and Medical Images

3.1. Introduction

This chapter will broadly cover the specific features of the signals and images most commonly used in clinical procedures. Of course, we may wonder if such analysis is essential when designing compression schemes dedicated to physiological signals and medical images. In other words, are these data so different from common signals and images (e.g. audio signals or images acquired by a digital camera)?

Physiological signals should not be considered as audio signals for the simple reason that they are provided by uncommon sources, having particular features. They differ from audio signals by their frequency content, by the correlation between samples and above all, by their dynamic. In fact, the quality of audio signals can be measured through regular auditory systems; this measurement procedure however cannot be applied to physiological signals. Consequently, an efficient coding device, in terms of compression rates and diagnostical quality, which will take into account the specific features of those signals, must therefore be employed.

Likewise, the specificities of medical images are found in the way that the meaning and hence the values of the pixels are represented. Most compression systems are tested on natural monochromatic or color images (e.g. the most commonly used image being the “Lena” image). The pixel intensity of those images corresponds to the reflection coefficient of natural light. ...