IT and predictive diagnosis

Abstract

Summary form only given. Information diminishes the uncertainty of a receiver to react in a certain optimum manner. It is a statistical property, or an event in a statistical population, and can be calculated by different entropy algorithms.. In image analysis one should distinguish between image content information (ICI), which is information that can be evaluated from the image itself and external knowledge which is not included in the image itself and which is commonly mandatory to interpret ICI in a correct manner. Diagnosis can then be defined as a mapping of external information on ICI. The specific algorithms how to do this will be discussed with focus on predictive diagnosis. Predictive diagnosis analyses intra- and extra-cellular pathways including gene abnormalities. It is the tool to develop and apply drug strategies in order to steer individualized cancer therapy. The significance of entropy measurements including structural (MST) entropy and image standardization is described in detail and demonstrated on individual cases.