AI models for personalised maternity care and ensure equitable access to maternal healthcare

Opportunity

Women from culturally and linguistically diverse (CALD) backgrounds encounter significant challenges when navigating the healthcare system. Language barriers, lower health literacy, and insufficient data expose them to a higher risk of receiving lower quality healthcare, inadequate service delivery, and poorer health outcomes compared to other Australians. The project aims to develop AI models for personalised maternity care that ensure equitable access by utilising diverse datasets to improve their applicability and effectiveness across different socio-ethnic groups.

This project will work on two overarching aims:

• Firstly, the project will recruit CALD women to collect maternity data specific to their experiences.
• Secondly, this data will be used to refine our current predictive models and develop new ones with the aim of providing
  accessible AI-driven early risk alerts and personalised treatment plans for women at risk of adverse birth outcomes.

This will be supported by expanding a comprehensive pregnancy knowledge database and creating a modelling framework that integrates clinical, lifestyle, environmental, and genetic data.

This approach will revolutionise the regular monitoring of pregnant women at risk of adverse delivery outcomes. By predicting risk alerts through linked data, the framework offers a strategy that can be extended to multiple pregnancy-related disorders. Adopting a person-centred care model will improve outcomes, save resources and reduce costs for care providers. There are potential downstream economic benefits from reducing inequities for CALD women, including improved workforce participation and reduced healthcare costs from chronic conditions tied to birth complications.

Project Objective

• Create the first ever stratified pregnancy risk evaluation of Australian CALD women and their underlying causes including detection of pathologies that can lead to adverse birth outcomes.
• Treatment decision based on personalised medical history, lifestyle, environment and/or genetic profiles.
• Address biases in predictive healthcare by developing models that are trained on heterogeneous datasets.
• Facilitate remote monitoring of maternal health through smart device integration, enabling clinicians to track vital health metrics in real time.