Data science for health management
Nearly 600,000 New Zealanders take medication for asthma and it is one of the most common causes of hospital admissions in children.
Alongside the devastating personal and community costs of asthma – as evidenced by the most recent New Zealand Health Information Service mortality data reporting more than 120 deaths within a year – the respiratory disease creates a huge economic burden.
Annual costs in New Zealand are estimated at over $1 billion, including more than $800 million from workdays lost, disability affected life years and mortality. Globally, asthma impacts more than 262 million people and in 2019 caused more than 455,000 deaths, according to the World Health Organization.
One of the critical ways to reduce the severity of poor asthma outcomes, including hospitalisations and mortality rates, is to understand which patients are most at risk and focus on preventive health management.
But how can you predict who is most at risk?
Researchers at AUT’s Data Science Research Centre are using data and machine learning models to find the answers.
One research project is using climatic parameters and air quality predictors as the input data streams, sourced from Auckland Council and NIWA, to predict the daily count of ED visits and admissions related to asthma – and ultimately help with health resource planning.
Another project, which Darsha Widana is working on as part of her PhD in computer and information sciences, uses multiple data sources to predict a personalised asthma risk control score for individual patients. This will provide data-based insights to help individuals and health practitioners decide on management and treatment plans.
The Director of the Data Science Research Centre, Associate Professor Farhaan Mirza, says data science is set to become one of the most intensively researched areas in computer science as it has massive potential applications.
“Data science can help us use data we already have to find answers and make accurate forecasts about the likelihood of events and outcomes in the future.”
Data Science Research Centre researchers are involved in a wide range of projects including the analysis of farm milk collection data to improve supply chain management, misinformation analysis of social media content, and designing systems to detect falls in infants and the elderly.