Entropy analysis in health informatics

Abstract

This chapter aims at presenting entropy measures that are now widely used in health informatics. Entropy measures have been introduced in 1990s and are derived from the theory of chaos. The traditional entropy-based analysis methods evaluate the degree of regularity of signals. We therefore present, first, the very well-known entropy measures used in the biomedical field: their theoretical background is detailed and some medical applications are presented. Then, more and very recent entropy measures are exposed and some of their used in health informatics are listed. This is performed from two points of view: for time series (uni-dimensional data) and for images (bi-dimensional data).