- विज्ञापन -
In a groundbreaking development, researchers from Australia’s Western Sydney University have introduced an artificial intelligence (AI)-driven tool that could redefine how Type 1 diabetes (T1D) is detected and treated. The innovation comes at a crucial time, as early and accurate diagnosis of T1D has remained a clinical challenge globally. The new tool, described in the journal Nature Medicine, uses a dynamic and responsive system known as the Dynamic Risk Score for Classification (DRS4C).
The new tool, described in the journal Nature Medicine, uses a dynamic and responsive system known as the Dynamic Risk Score for Classification (DRS4C). Unlike conventional genetic testing methods, which provide a static view of lifetime risk, this AI-based score offers real-time insights into a person’s current
likelihood of developing T1D.
Unlike conventional genetic testing methods, which provide a static view of lifetime risk, this AI-based score offers real-time insights into a person’s current likelihood of developing T1D. The DRS4C model is based on microRNAs—tiny RNA molecules found in the blood that can reflect changes in the body’s biological state.
By measuring these molecules and feeding the data into an AI framework, researchers can track how an individual’s risk of T1D fluctuates over time and predict who might benefit from early intervention.“Type 1 diabetes that begins early—particularly before age 10—is known to be more aggressive and can reduce life expectancy by up to 16 years,” said Professor Anand Hardikar, the lead investigator of the study.
“Having a predictive model like DRS4C allows clinicians to intervene earlier and more effectively.”To build this risk assessment tool, the team analysed an extensive dataset comprising 5,983 samples from study participants in seven countries, including India, Australia, Canada, the US, Denmark, Hong Kong, and New Zealand.
They later validated the model on a separate group of 662 individuals. Remarkably, the risk score could predict— within just an hour after therapy—which individuals would remain insulin-free, suggesting potential in tailoring treatment strategies. What sets this model apart is its potential to do more than just predict risk.
Researchers found that the score could also indicate how well a person would respond to emerging therapies that aim to delay the progression of T1D. Additionally, it showed promise in distinguishing between Type 1 and Type 2 diabetes—two conditions that often present similarly in early stages but require very different treatment approaches.Dr. Mugdha Joglekar, co-lead researcher, emphasized the significance of using dynamic risk markers over traditional genetic indicators.
“Think of genetic testing as knowing you live in a flood zone—it tells you there’s a long-term risk. But DRS4C acts like a waterlevel gauge, giving real-time updates that help doctors make timely decisions,” she said. This, she added, also helps reduce stigma by allowing for adaptive, rather than deterministic, health monitoring.As the prevalence of diabetes continues to rise globally, innovations like DRS4C could prove pivotal in personalizing care, optimizing therapies, and improving longterm outcomes for those at risk of or living with Type 1 diabetes.
