Resumo

RAMOS, Esmeralda; NUNEZ, Haydemar  e  CASANAS. Designing an agent-based system to support the diagnosis of human semen quality. Rev. Fac. Ing. UCV [online]. 2011, vol.26, n.3, pp.15-25. ISSN 0798-4065.

The analysis of human semen quality is a complex process that requires professionals to have a wide range of knowledge and great expertise. In order to support them in this task, in this paper an design of an agent-based system is proposed. The agents are able to: a) Classify the ejaculate quality using a fuzzy hybrid model of case-based reasoning and rulebased reasoning, b) Manage patients’ medical histories, c) Retrieve specialized information from a domain ontology of Andrology, and d) Analyze the morphology of sperm heads from digitized images using image processing techniques and machine learning techniques (neural networks and support vector machines). In addition, the ontology developed is the first version available for this domain. The system was design with the INGENIAS methodology using a Web-oriented approach.

Palavras-chave : Agent; Ontologies; INGENIAS; Case-based reasoning; Neural networks; Human semen.