2019 – Data Mining and Risk Analysis Supporting Decision in Brazilian Public Health System

Date of Conference: 14-16 Oct. 2019
Date Added to IEEE Xplore27 February 2020
 ISBN Information:
Publisher: IEEE
Conference Location: Bogota, Colombia, Colombia

Abstract:

Health data monitoring is a crucial activity to reduce maternal, neonatal and infant mortality rates. Available data in Brazilian health databases point that It is possible to predict death risk in the early stages of gestation and infant development. In this research, we consider the information availability still in the gestational period to propose different death risk prediction models for this public of interest. We also detail the data mining process to apply machine learning-based techniques in death risk classification for maternal, neonatal and infant patients. We present an experiment pipeline to estimate average performance and evaluated machine learning models with different features combinations. Additionally, we show a web service which provides multiple predictive models by information availability. Results show Random Forest obtaining better performance when compared to the other machine learning methods.

Publicado em 2. Artigos CIENTÍFICOS | Deixe um comentário

2019 – GIRLS, a Gateway for Interoperability of Electronic Health Record in low-cost System

Date of Conference: 14-16 Oct. 2019
Date Added to IEEE Xplore27 February 2020
 ISBN Information:
Publisher: IEEE
Conference Location: Bogota, Colombia, Colombia
Abstract:

Clinical information about patients should be consistent, complete and available to health professionals, ensuring quality care. This information is recorded on paper or on electronically-stored in a digital format. A few years ago, both Brazilian government and private healthcare providers invested in Information and Communication Technology to replace paper medical record with Electronic Medical Record (EMR). Nowadays, EMRs are evolving into Electronic Health Record (EHR), which allows for interoperability between different systems. While brazilian public health system recommends the OpenEHR standard as an information model for EHR, private providers in Brazil have adopted the HL7 FHIR standard. This article proposes the GIRLS, a low-cost gateway for EHR interoperability that uses both standards. As proof of concept, a chikungunya OpenEHR archetype and an equivalent FHIR feature have been implemented. This archetype is available to the Clinical Knowledge Manager (CKM), the largest online repository of archetypes on the Web.
Publicado em 2. Artigos CIENTÍFICOS | Deixe um comentário

2019 – Quality of Health Service, Optimization an IoT Solution wit Diffserv and Protocols

Date of Conference: 14-16 Oct. 2019
Date Added to IEEE Xplore27 February 2020
 ISBN Information:
Publisher: IEEE
Conference Location: Bogota, Colombia, Colombia

Abstract:

This work presents the Quality of Health Service (QhS), an IoT solution for a health patient monitoring environment and proposes an optimization mechanism with the Diffserv and EWS protocols. A mobile application is implemented for the specific healthcare team to have access to the system for viewing and modifying patient information. The analysis of vital signs in QhS solution took into consideration the network paradigm and IoT service, as well as the risk of the patient based on the EWS protocol. Thus, the QhS combines the Diffserv (network level) and EWS (application level) protocols for the optimization of data traffic in the system monitoring information and alerts. Additionally, It results in energy saving, still a vital resource in IoT devices.
Publicado em 2. Artigos CIENTÍFICOS | Deixe um comentário

2019 – LARIISA: an intelligent platform to help decision makers in the brazilian health public system

LARIISA: an intelligent platform to help decision makers in the brazilian health public system

Publication:WebMedia ’19: Proceedings of the 25th Brazillian Symposium on Multimedia and the WebPages 501–504https://doi.org/10.1145/3323503.3362122

ABSTRACT

LARIISA is an intelligent framework for decision-making in public health systems. The project had its initial ideas conceived in 2009. Since then it has evolved in the academic and market perspective, becoming a product in 2018 called GISSA. This article presents the architectural evolution of LARIISA, the functionalities implemented, the scientific and commercial results achieved with GISSA. Ontology and Data Mining (DM) are technologies that support their inference mechanisms. A semantic portal is proposed for GISSA and a DM application is presented.

Publicado em 2. Artigos CIENTÍFICOS | Deixe um comentário

2018 – MARCIA: Applied Clinical Record Management : Eletronic Health Record Applied with EHRServer

MARCIA: Applied Clinical Record Management : Eletronic Health Record Applied with EHRServer
Date of Conference: 17-20 Sept. 2018
Date Added to IEEE Xplore12 November 2018
 ISBN Information:
INSPEC Accession Number: 18233025
Publisher: IEEE
Conference Location: Ostrava, Czech Republic

Abstract:

The quality of services provided to patients in the health area is directly related to the quality of clinical information. In addition, this information must be consistent, secure and available to health professionals, even though health data is usually distributed across heterogeneous systems. The Electronic Patient Record (EPR) was proposed and applied to minimize integration problems through the construction of health information systems. This work proposes a methodology for the development of interoperable and flexible systems, using the EHRServer framework of the OpenEHR standard. As a case study, this methodology has been applied in Aracati/CE since March / 2017, in the context of the Chikungunya disease. The methodology is supported by a system that implements a set of OpenEHR archetypes representing the clinical treatment of Chikungunya. The system was tested in a Basic Health Unit. The archetypes and the MARCIA Templates were made available to the Clinical Knowledge Manager (CKM), the largest online repository of archetypes on the Web.
Publicado em 2. Artigos CIENTÍFICOS | Deixe um comentário

2017 – Predicting Hypertensive Disorders… Using the Random Forest Approach

Date of Conference: 21-25 May 2017
Date Added to IEEE Xplore31 July 2017
 ISBN Information:
Electronic ISSN: 1938-1883
INSPEC Accession Number: 17065434
Publisher: IEEE
Conference Location: Paris, France

Abstract:

The incidence of hypertension associated with pregnancy contributes significantly to increase maternal and fetal deaths during pregnancy and childbirth. Due to its high incidence rate and several complications, the study of this disorder has been subject of numerous investigations in an attempt to determine its prevention and improve the treatment conduction. In this context, this paper uses a data mining (DM) technique, named random forest (RF), applied to health care to early identification of these disorders. It also presents the modeling, performance assessment, and comparison with other DM methods to evaluate the performance of the proposed model. Results showed that the RF classifier had a regular performance, presenting the best values for true positive Rate (TP Rate) and recall in the prediction of preeclampsia superimposed on chronic hypertension compared to the other experimented classifiers. Even finding a good performance to predict hypertensive disorders, other tree-based methods need to be evaluated, as well as other DM techniques. Discovering reliable information of pregnant women suffering from the hypertensive disease is an important path to reduce the high rate of deaths, mainly, in developing countries where 99% of these deaths occur.
Publicado em 2. Artigos CIENTÍFICOS | Deixe um comentário

2017 – Using Linked Data in the Data Integration for Maternal and Infant Death Risk of the SUS in the GISSA Project 

Using Linked Data in the Data Integration for Maternal and Infant Death Risk of the SUS in the GISSA Project 

  • Published in: WebMedia – XVI Workshop de Ferramentas Aplicações  (17-20 October)
  • DOI: 10.1145/3126858.3131606 –  Conf Location: Gramado, RGS – Brazil 
  • ACM New York, NY, USA ©2017 –  table of contents ISBN: 978-1-4503-5096-9

ABSTRACT

Making good governance decisions is a constant challenge for Public Health administration. Health managers need to make data analysis in order to identify several health problems. In Brazil, these data are made available by DATASUS. Generally, they are stored in distinct and heterogeneous databases. TheLinked Data approach allow a homogenized view of the data as a unique basis. This article proposes a ontology-based model andLinked Data to integrate datasets and calculate the probability of maternal and infant death risk in order to give support in decision-making in the GISSA project.

Publicado em 2. Artigos CIENTÍFICOS | Deixe um comentário

2017 – Using Predictive Classifiers to Prevent Infant Mortality in the Brazilian Northeast

Date of Conference: 12-15 Oct. 2017
Date Added to IEEE Xplore18 December 2017
 ISBN Information:
INSPEC Accession Number: 17433294
Publisher: IEEE
Conference Location: Dalian, China

Abstract:

Despite the fact that infant mortality rates have been decreased in recent years, this issue stills being considered alarming to Brazilian health system indicators. In this context, the GISSA framework, an intelligent governance framework for Brazilian health system, emerges as a smart system for the Federal Government program, called Stork Network. Its main objective is to improve the healthcare for pregnant women as well as their newborns. This application aims to generate alerts focusing on the health status verification of newborns and pregnant woman to support decision-makers in preventive actions that may mitigate severe problems. Therefore, this paper presents the LAIS, an Intelligent health analysis system that uses data mining (DM) to generate newborns death risk alerts through probability-based methods. Results show that the Naïve Bayes classifier presents better performance than the other DM approaches to the used pregnancy data set analysis of this work. This approach performed an accuracy of 0.982 and a Receiver Operating Characteristic (ROC) Area of 0.921. Both indicators suggest the proposed model may contribute to the reduction of maternal and fetal deaths.
Publicado em 2. Artigos CIENTÍFICOS | Deixe um comentário

2019 – A visualization and analysis approach of cyclist data obtained through sensors  (*) Best Paper !

Date of Conference: 6-11 Aug. 2017
Date Added to IEEE Xplore22 October 2018
 ISBN Information:
INSPEC Accession Number: 18168104
Publisher: IEEE
Conference Location: Natal, Brazil

Abstract:

Solutions for smart cities are being created everywhere in the world, using technology to improve urban infrastructure and make urban centers more efficient and better to live. In this way, the proposal of this work is based on the capture of information through sensors in a smart city context. Sensors are coupled on a bicycle and connected to an Arduino and a Mobile Application. After this, the captured data are saved in a cloud database, displayed and analyzed through a Web Application. In this work, our methodology is organized in three main phases: (i) data collection of the surface in which the cyclist is traveling, and the ultrasonic distance sensor, to identify areas of risk based on the proximity of objects from bicycle, (ii) data analysis and data classification, using machine learning concepts and (iii) data visualization, using map views in a Web Application. This methodology allows the identification of injury risk situations to cyclists. The main contributions of this work are surfaces classification with data collected by the accelerometer and ultrasonic sensor generating useful information through simple data. Real experiments were conducted at Fortaleza (Ceara, Brazil) and Aracati (Ceara, Brazil). This work brings new perspectives to collaborative data collection for identification of injury risk situations to cyclists, since it can be used to suggest routes based on these risk indicators and offer a secure environment for cyclists.
Publicado em 2. Artigos CIENTÍFICOS | Deixe um comentário

2016 – Performance Evaluation od Predictive Classifiers for Pregnancy Care 

Performance Evaluation od Predictive Classifiers for Pregnancy Care

Date of Conference: 4-8 Dec. 2016
Date Added to IEEE Xplore06 February 2017
 ISBN Information:
INSPEC Accession Number: 16654639
Publisher: IEEE
Conference Location: Washington, DC, USA

Abstract:

Hypertensive disorders are the leading cause of deaths during pregnancy. Risk pregnancy accompaniment is essential to reduce these complications. Decision support systems (DSS) are important tools to patients’ accompaniment. These systems provide relevant information to health experts about clinical condition of the patient anywhere and anytime. In this paper, a model that uses the Naive Bayesian classifier is introduced and its performance is evaluated in comparison with the Data Mining (DM) classifier named J48 Decision Tree. This study includes the modeling, performance evaluation, and comparison between models that could be used to assess pregnancy complications. Evaluation analysis of the results is performed through the use of Confusion Matrix indicators. The founded results show that J48 decision tree classifier performs better for almost all the used indicators, confirming its promising accuracy for identifying hypertensive disorders on pregnancy.
Publicado em 2. Artigos CIENTÍFICOS | Deixe um comentário