Here are some things we’re working on right now. Under each project
description, you can find the solutions developed using Artificial
Intelligence, Geographic Information Systems (GIS), and data
visualization.
Healthcare Data Analytics
Data-driven risk stratification for preterm birth in Brazil: development of a machine learning-based innovation for health care
Preterm birth (PTB) is a growing health issue worldwide, currently
considered the leading cause of newborn deaths, accounting for 1.09
million deaths/year of children under the age of 5. In Brazil
specifically, PTB rates doubled from 6.5% in 2004 to 12% in 2013.
Identifying the preventable causes and performing early risk
stratification of pregnant women has been shown to be effective in
developing preventive strategies aiming to reduce PTB. However, although
data exists, the ability to identify at-risk pregnancies and to enroll
women in prevention strategies has been weakened by the complexity of
the associated risk factors. PTBs involves several components, such as
mothers? age higher than 35, primiparity, twin delivery, hypertension,
diabetes, previous preterm delivery, among other elements. Our goal is
to combine different national-level data sources to understand the main
predictors of PTB and develop a machine-learning-based predictive model
to conduct automated risk stratification for PTB in Brazil at the point
of care level, integrated with advanced data visualization for clinical
decision support.
Live birth monitoring dashboard
Geospatial Artificial Intelligence-based microplanning for health
campaigns in low resource settings: A solution to overcome COVID-19
challenges
The year 2020 is challenging the health systems across the globe. The
Covid-19 pandemic is pressing researchers, policymakers, and
researchers to develop new strategies to overcome the new and old
challenges. Usually, campaign-based delivery is the main driver of
health systems to leverage the impact of interventions aiming to achieve
disease control. Despite the importance of health campaigns very often
the results obtained are detached from the goals initially defined.
Nearly all public health outreach efforts, from vaccinations to bed nets
to HIV treatment, depend on accurate target population denominators to
estimate resource needs and project costs, as well as to measure and
assess results and impacts. The impossibility to identify the
location and distribution of the eligible population to be reached
arises as a major performance challenge regarding health campaigns.
A microplan is the tool usually used to define a population-based set
of components for delivering health-care interventions. Although its
importance there are few innovations dedicated to improving the quality,
reduce the time needed, and the amount of effort necessary to develop
strong microplanning strategies for health campaigns. Considering
the challenges highlighted above the objective of the project is to
develop an open-source platform capable to create customized
microplanning for health campaigns using satellite imagery to estimate
the up to date population eligible to be reached.
Development and implementation of a culturally relevant machine
learning decision support tool to improve risk stratification of
traumatic brain injury patients in Tanzania and Uganda
Every year, an estimated 69 million people sustain a traumatic brain
injury (TBI), 10 million of which result in hospitalization or death. A
disproportionately high burden of this injury occurs in low and
middle-income countries (LMICs) where shortages of diagnostic
technologies and highly-skilled providers are exposed by high patient
volumes. The resource limitations of health care facilities in LMICs
contribute to the treatment and diagnostic delays which are known
contributors to poor TBI outcomes. The use of prognostic modeling as a
clinical decision support tool could optimize the use of existing
resources and support timely treatment decisions in LMICs. Our long-term
goal is to create a machine-learning-based decision support tool to
help optimize clinical decision making for TBI care for low-resource
settings across different levels of care. The objective of this
application is to expand our efforts from Tanzania and translate, adapt,
implement and evaluate the feasibility of this technology in clinical
decision making in Tanzania, Uganda, India, and Pakistan with the
following specific aims.
Artificial intelligence risk calculator of Traumatic Brain Injury outcome
Application of eHealth technologies for disease prevention
The goal of this project is to integrate as well as develop eHealth
related technologies for the prevention of acute episodes related to
chronic conditions. The use of advanced techniques of data analysis such
as artificial intelligence, data mining, BIG DATA, machine learning,
and natural language processing will be applied for the identification
of parameters responsible for chronic diseases. Once the predictive
models associated with the diseases of interest have been designed,
strategies for follow-up, monitoring, and intervention will be defined
with the aim of preventing acute episodes.
Health performance assessment
Program for Improving Access and Quality of Primary Care – PMAQ
The National Program for Improving Access and Quality of Primary Care
is a program that seeks to induce processes capable to increase the
capacity of federal, state and municipal administrations, as well as
Primary Care Teams, to offer services that ensure greater access and
according to the specific needs of the population. The program seeks to
induce the expansion of access and improvement of the quality of basic
health care, with a guarantee of a comparable quality standard
nationally, regionally, and locally in order to allow greater
transparency and effectiveness of government actions directed to Primary
Health Care throughout the country.
Mobile Health (mHealth)
Identification of global trends in mhealth, aiming to prevent accidents and work-related diseases
The Unit of Studies and Prospects (UNIEPRO), in compliance with its
objective of Generating knowledge to support positioning, strengthen
synergy, identify current and future opportunities and threats for SESI,
SENAI, and IEL, established for the year 2017 the Convergence Project
MHEALTH. The use of mobile and wireless technologies to support the
attainment of health objectives has the potential to transform health
care delivery around the world. A powerful combination of factors is
driving this change. These include rapid advances in mobile technologies
and applications, an increase in new opportunities for integrating
mobile health into existing e-health services, and the continued growth
of mobile cellular network coverage. The objectives of the project are
to establish the state of the art, identify and describe the main
characteristics of mobile health technologies (mHealth) that can add
value and contribute to the reduction of absenteeism in the industry
workers through the prevention of accidents and diseases, as well as
promoting health and a healthy lifestyle; to map trends in these
technologies that could be introduced on the world market by 2030;
identify and characterize the main users of these technologies; identify
the major worldwide Research Centers that are already using mHealth
technologies and their modes of use as well as the features and
applications of Big Data involved; identify the types of competencies of
the multidisciplinary teams involved in the development and use of
these technologies by these Centers.
Geographic Information Systems – GIS
Geospatial model for risk of preterm births in Brazil
The main objective of this project was to identify the factors that
influence premature birth, considering its spatial distribution in
Brazil. Method: A model based on data mining techniques was elaborated
to obtain the variables that can predict the weighted prematurity
patterns by geospatial location. The statistical analysis will be
performed from a Bayesian spatial model, the variable of interest being
the number of preterm births in the municipality. The Poisson
distribution will be adopted considering the expected number of preterm
births in a given municipality and the relative risk of that
municipality. Sociodemographic variables will be considered and related
to the quality of health services offered. In the end, the result will
be a panel model showing how spatial distribution affects premature
birth in each Brazilian municipality. Expected results: It is expected
that from the conception of a monitoring system of health outcomes by a
geospatial approach, it will be expected to subsidize quality
improvement and fair access to services offered, especially in primary
health care from regular adoption and systematized models of geospatial
risk by the family health strategy and health service managers for
preterm birth prevention.