Automated Diagnosis Through Ontologies and Logical Descriptions: The ADONIS Approach

The combination of the burgeoning interest in efficient and reliable Health Systems and the advent of the Information Age represent both a challenge and an opportunity for new paradigms and cutting-edge technologies reaching a certain degree of maturity. Hence, the use of Semantic Technologies for Automated Diagnosis could leverage the potential of current solutions by providing inference-based knowledge and support on decision-making. This paper presents the ADONIS approach, which harnesses the use of ontologies and the underlying logical mechanisms to automate diagnosis and provide significant quality results in its evaluation on real-world data scenarios. medical knowledge (Liu et al., 2009). Since many artificial intelligence approaches have dealt with the diagnosis problem and its application in complex environments such as medical domains (Fuentes-Lorenzo et al., 2009), semantic technologies can provide a strong cutting-edge baseline for knowledge-oriented medical diagnosis systems. The semantic DOI: 10.4018/jdsst.2011010102 22 International Journal of Decision Support System Technology, 3(1), 21-39, January-March 2011 Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. technologies (Berners-Lee et al., 2001), which have been developed and improved alongside the advancement of the Semantic Web, can be exploited to reveal machine-readable latent relationships within specific diagnostic-related information in the medical discipline, where the homogeneity of terminology is particularly problematic (Fuentes-Lorenzo et al., 2009). The specification of the domain knowledge (in our case, the medical domain) by the use of ontologies (Fensel, 2002) provides a knowledge-based system with the opportunity of adding the semantics of the domain, stating an explicit conceptual description of the domain (Fuentes-Lorenzo et al., 2009). However, the efficiency and soundness of Semantic descriptions must be backed up by their underlying logic. The lattice of logic languages and formalisms is not a trivial issue and hence, an ontology must be perfectly defined and explained to serve as a basis for real-world medical applications. For this reason, an accurate and checked ontology should be defined in order to create a base for the medical diagnosis systems. Also, the description of the diseases, symptoms, laboratory tests and other clinical parameters should be done with rigor and checked by medical doctors. This description is the problem presented in the most of actual software for clinical diagnosis where not all the possibilities are taken into account, because in some cases this software are not able to make the correct inference of the disease. This paper presents ADONIS, an architecture that includes a well-structured ontology for automated diagnosis and a three-fold formalization based on Description Logic that will allow to future medical systems that implements these techniques to perform real and more accurate diagnosis that traditional and actual systems. The remainder of the paper is organized as follows. Section 2 outlines related research in the area. In Section 3, the architecture for the ADONIS approach is presented, which demonstrates technological support and potential solutions for problems in current medical diagnosis based in semantics and logical descriptions. An evaluation using calculation of precision and recall rates of the system is also presented in section 4. Conclusions and future work are discussed in Section 5.