List of results published directly linked with the projects co-funded by the Spanish Ministry of Economy and Competitiveness under the María de Maeztu Units of Excellence Program (MDM-2015-0502).

List of publications acknowledging the funding in Scopus.

The record for each publication will include access to postprints (following the Open Access policy of the program), as well as datasets and software used. Ongoing work with UPF Library and Informatics will improve the interface and automation of the retrieval of this information soon.

The MdM Strategic Research Program has its own community in Zenodo for material available in this repository   as well as at the UPF e-repository   

 

 

Back Bill and Melinda Gates Foundation support research on machine learning for high-risk pregnancies in Pakistan

Bill and Melinda Gates Foundation support research on machine learning for high-risk pregnancies in Pakistan

The Bill and Melinda Gates Foundation has recently awarded the project “Machine learning from fetal flow waveforms to predict adverse perinatal outcomes”, led by Aga Khan University (AKU) in Karachi (Pakistan). Pakistan is one of the countries where stillbirth rate and early neonatal mortality rates are among the highest in the world. The project aims to prove that a machine learning approach can contribute to predict adverse perinatal and neonatal outcomes.

18.12.2017

 

The Bill and Melinda Gates Foundation has recently awarded the project “Machine learning from fetal flow waveforms to predict adverse perinatal outcomes”, led by Aga Khan University (AKU) in Karachi (Pakistan).

Pakistan is one of the countries where stillbirth rate and early neonatal mortality rates are among the highest in the world. The aim of this study is to get a proof of concept for using a computational model of fetal haemodynamics, combined with machine learning based on  Doppler patterns of the fetal cardiovascular, cerebral and placental flows, to predict identify those at increased risk of adverse perinatal outcomes such as stillbirth, perinatal mortality and other neonatal morbidities.

At-risk and high-risk pregnancies can be detected using ultrasound, biochemical screening and genetic analysis and maybe offered an early delivery if the baby should show signs of difficulty. However, poor infrastructure and lack of resources as well as training limit the availability of these tools to detect high risk and at-risk pregnancies. The project aims to prove that a machine learning approach based on Doppler flow patterns of the fetus and placenta can be used to predict adverse perinatal and neonatal outcomes like still birth, prematurity, IUGR and early neonatal mortality.

Bart Bijnens, from the PhySense research group, and Gemma Piella, from the SIMBIOsys research group, both part of the Barcelona MedTech Unit in DTIC, will be responsible for data analysis. Analysis will include developing the machine learning algorithm required to predict adverse perinatal outcomes. To fuse and order multivariate heterogenous data, and be agnostic of any outcome, the Multiple Kernel Learning (MKL) methodology will be utilized in a non-supervised setting. The data would undergo non-linear dimensionality reduction to extract the most discriminate features from the multiple and rich input variables allowing for independent complex interrelation with abnormalities. Identifying clusters of individuals, after identifying the most relevant discriminate patterns in the population, and relating this to clinical findings and outcome, subsequently allows for the identification of phenogroups with similar outcome. When data from a new individual, not used for the learning phase, is available, the most similar patients from the learning phase can be identified (by projection of input data into the learned low-dimensional space) and the pheno-group membership can be quantified from which predicted outcome can be derived assuming it will be similar to the expected outcome of the group.

 

Multimedia

Categories:

SDG - Sustainable Development Goals:

Els ODS a la UPF

Contact