A Preeclampsia Diagnosis Approach using Bayesian Networks Mário W. L. Moreira 1,2 , Joel J. P. C. Rodrigues 1,3 , Antonio M. B. Oliveira 2 , Ronaldo F. Ramos 2 , Kashif Saleem 4 1 Instituto de Telecomunicações, University of Beira Interior, Portugal 2 Instituto Federal do Ceará, Brazil 3 University of Fortaleza (UNIFOR), Ceará, Brazil 4 Center of Excellence in Information Assurance (CoEIA), King Saud University (KSU), Riyadh, Kingdom of Saudi Arabia mario.moreira@it.ubi.pt; joeljr@ieee.org; amauroboliveira@gmail.com; ronaldo@ifce.edu.br; ksaleem@ksu.edu.sa Abstract— Hypertension is the main cause of maternal death. Preeclampsia can affect pregnant women before or during pregnancy. Identification of patients with higher risk for preeclampsia allows some precautions that are taken to prevent its severe disease and subsequent complications. In medicine, there are different situations that deal with a large range of information, which needs a thorough assessment to be able to help experts in the decision-making process. Smart decision support systems allow grouping all existing information and finding pertinent information from it. Bayesian networks offer models that allow the information capture and handle situations of uncertainty. This paper proposes the construction of a system to support intelligent decision applied to the diagnosis of preeclampsia using Bayesian networks to help experts in the pregnant’s care. The processes of qualitative and quantitative modeling to the construction of a network are also presented. The main contribution of this work includes the presentation of a Bayesian network built to help decision makers in moments of uncertainty in care of pregnant women. Keywords— Decision support systems; Bayesian networks; Pregnancy; Hypertencion; Modeling I. INTRODUCTION The quality of care provided by health services for pregnant women is the most important action to control maternal mortality. Access to these services presents a major impact on these reductions, since they have enough quality to identify risks. Information systems offer the ability to monitor and evaluate the pregnant health notifying the experts in good time about a given complication that can occur during pregnancy. This allows the physicians make better decisions on diagnosis and establish better medical procedures and treatment. The most frequently diseases during pregnancy are infectious, especially those that reach the urinary tract. These diseases can cause severe complications like increasing the risk of miscarriage and anticipation of the birth labor. However, the main concerns of obstetricians are related to metabolic syndromes such as preeclampsia and gestational diabetes, which are more fatal for both mothers and babies. Preeclampsia occurs when a pregnant woman has high blood pressure (above 140/90 mmHg) at any time after the 20 th pregnancy week and disappears before 12 weeks postpartum [1]. Besides high blood pressure, other complications such as excessive protein in the urine should occur in a diagnosis of preeclampsia. In [2], the authors identify high-risk pregnancy complaints and propose a method that analyses the Doppler signal to identify these conditions. Conclusions show that complications with pregnancy are associated with hypertensive disorders (preeclampsia), intra-uterine growth restriction of fetus, and gestational diabetes mellitus. In [3], the importance of a reliable diagnosis with accurate measurement of blood pressure and proteinuria is discussed. The authors also present cases where preeclampsia goes undiagnosed due to lack of appropriate equipment and limited resources in laboratories. Cheng et al. [4] analyze the relationship between the blood pressure and risk factors of pregnant women. To determine the effects over gestational age a varying coefficient model was established. Results show that effects of known risk factors change with gestational age. This result is an important knowledge for understanding the causes of gestational hypertension. In [5] the authors investigate the alterations in the progress of a normal pregnancy and pregnancy disorders associated with hypertension. This research work uses the Joint Symbolic Dynamics method. Mukherjee et al. [6] use Discriminant Analysis and k-means clustering to predict preeclampsia based on lipid parameters. This technique is used to separate the pregnant women in two groups named preeclampsia and control, so that a new patient can be classified into any of these groups according to estimated values of the parameters. In this context, several efforts have been performed in order to develop a system for giving support to experts in pregnant’s care. Then, this work aims to investigate how Bayesian networks can support clinicians to identify high-risk pregnancy. It proposes the use of a statistical model based on Bayesian networks to better classify the seriousness of a problem helping the decision-makers in uncertainty moments. These systems applied to healthcare offer the possibility to monitor and evaluate pregnant’s health and notify any complications that can occur during the pregnancy in a due time. The main contribution of this work includes a proposal of a smart system based on Bayesian networks to support decision makers in pregnant women monitoring using the Noisy-or classifier.