Key points from article :
Respiratory diseases like asthma and chronic obstructive pulmonary disease (COPD) affect millions worldwide, yet they remain difficult to treat and diagnose precisely. Roan Zaied, whose own childhood was marked by severe asthma attacks, has dedicated her research career to tackling this problem. While asthma is highly diverse—ranging from childhood onset to exercise-induced forms—most patients are still given the same standard treatments. This one-size-fits-all approach doesn’t work for everyone, with 5–10% of people failing to respond to existing medications. Zaied’s early work at the University of Auckland’s Liggins Institute focused on genetics, with the aim of identifying specific genes that could help tailor therapies to different asthma subtypes, a step toward personalised medicine.
Her research has since shifted toward COPD, a progressive lung disease that is the fourth leading cause of death worldwide and affects around 15% of New Zealanders over 40. COPD is usually diagnosed with spirometry, a breathing test that only detects the disease after irreversible lung damage has already occurred. Because there is no cure, earlier diagnosis could make a critical difference. Zaied’s goal is to create tools that identify people most at risk before major damage sets in, opening the door to prevention and better-targeted care.
To achieve this, Zaied is harnessing the power of machine learning—an artificial intelligence technique capable of detecting complex patterns in large datasets. She analysed information from 10,000 smokers tracked over five years, combining genetic profiles, CT scans, and clinical risk factors like age, sex, and smoking history. Her AI model was able to predict COPD with 86% accuracy, identifying subtle lung changes and risk markers invisible to traditional tests.
The next step is to test whether the model can prospectively predict who will go on to develop COPD, potentially transforming how doctors approach prevention and diagnosis. If successful, Zaied’s work could mark a major leap forward in personalised respiratory medicine—turning raw data into tools that help patients receive earlier, more precise, and more effective care.
