The IRMA Community
Newsletters
Research IRM
Click a keyword to search titles using our InfoSci-OnDemand powered search:
|
A Complex Data Warehouse for Personalized, Anticipative Medicine
Abstract
With the growing use of new technologies, healthcare is nowadays undergoing significant changes. The development of electronic health records shall indeed help enforcing personalized, lifetime healthcare and pre-symptomatic treatment, as well as various analyses over a given population of patients. Such an information-based medicine has to exploit medical decision-support systems and requires the analysis of various, heterogeneous data, such as patient records, medical images, biological analysis results, etc. (Saad, 2004). In this context, we work with a physician on large amounts of data relating to high-level athletes. His aim is to make the analyzed subjects the managers of their own health capital, by issuing recommendations regarding, e.g., life style, nutrition, or physical activity. This is meant to support personalized, anticipative medicine. To achieve this goal, a decision-support system must allow transverse analyses of a given population and the storage of global medical data such as biometrical, biological, cardio-vascular, clinical, and psychological data. Data warehousing technologies are now considered mature and can form the base of such a decision-support system. Though they primarily allow the analysis of numerical data, the concepts of data warehousing remain valid for what we term complex data. In this context, the warehouse measures, though not necessarily numerical, remain the indicators for analysis, and analysis is still performed following different perspectives represented by dimensions. Large data volumes and their dating are other arguments in favor of this approach. Data warehousing can also support various types of analysis, such as statistical reporting, on-line analysis (OLAP) and data mining. In this paper, we present the design of the complex data warehouse relating to high-level athletes. It is original in two ways. First, it is aimed at storing complex medical data. Second, it is designed to allow innovative and quite different kinds of analyses to support: 1. personalized and anticipative medicine (in opposition to curative medicine) for well-identified patients; 2. broad-band statistical studies over a given population of patients. Furthermore, the system includes data relating to several medical fields. It is also designed to be evolutionary to take into account future advances in medical research.
|
|