A Fuzzy Intelligent Agent for Analysis and Classification of Fetuses' Cardiac Signals

The signal analysis provided by fetal monitoring equipment called cardiotocograph such as fetal heart rate, allows early diagnosis about possible health problems related to the fetus and therefore the safety of their well being. This paper presents a proposal for an intelligent system that, after analyzing the signals collected by cardiotocograph automatically provides the diagnosis. The analysis is done through a hazy inference system (fuzzy) modeled in partnership with the medical staff of the Maternity School Assis Chateaubriand (MEAC). This system is implemented using the Matlab software. Tests are performed, obtaining promising results according to team MEAC since achieves a hit percentage from 85% to 93% in the analysis carried out in two databases. It is important to note that this work has a multi-institutional partnership between Brazil and Germany, involving the Federal University of Ceará (UFC), through the Department of Computer Engineering (DETI) and the Maternity School Assis Chateaubriand (MEAC), Technische Universität München (TUM), the Bundeswehr Universität München and the company Trium Analysis Online GmbH.