The Doppler Project: Machine learning for high-risk pregnancies in Pakistan

TitleFetal Doppler for Antenatal Risk Stratification (Doppler project)

TeamZahra Hoodbhoy (AKU), Babar Hasan (SIUT), Shazia Mohsin (SIUT), Devyani Chowdhury (Cardiology Care for Children, USA), Sergio Sanchez (Universitat Pompeu Fabra​), Josa Pratz (UPF), Bart Bijnens (UPF)

Objectives: Globally, Pakistan has the highest rate of stillbirths and early neonatal mortality (death within 1 week of life). Flow changes in several major arteries of the feto-placental circulation, as detected by Doppler echocardiography, may provide important information regarding fetal compromise which may lead to perinatal morbidity and mortality. 

The objective of this project is to build a machine learning algorithm on maternal and fetal characteristics along with Doppler waveforms to predict an adverse perinatal outcome.

Sites: Rehri Goth and Ibrahim Hyderi, Karachi
Timeline: Work is ongoing; anticipated completion, October 2024

SponsorBill and Melinda Gates Foundation $2,525,000​

A 1:16 min video that summarises the Doppler project, prepared for dissemination. Credit: Rahim Sajwani/AKU Department of Paediatrics & Child Health

Dissemination, presentations:  

Hoodbhoy Z, Hasan B, Jehan F, Bijnens B, Chowdhury D. Machine learning from fetal flow waveforms to predict adverse perinatal outcomes: a study protocol. Gates open research. 2018;2.

Naz S, Hoodbhoy Z, Jaffar A, Kaleem S, Hasan BS, Chowdhury D, Gladstone M. Neurodevelopment assessment of small for gestational age children in a community-based cohort from Pakistan. Archives of disease in childhood. 2023 Apr 1;108(4):258-63.
Presentation and Panel Discussion at Grand Challenges Annual Meeting in Dakar, Senegal 2023