AI technology revolutionising medical research process
In a significant breakthrough for healthcare research, UQ HERA Fellows, Professor Guido Zuccon, Dr Teerapong Leelanupab and Associate Professor Bevan Koopman are developing a new artificial intelligence platform to transform how medical evidence is gathered and analysed. This innovation addresses one of the most pressing challenges in evidence-based medicine: the time-consuming process of systematic reviews.
The Challenge of Keeping Pace with Medical Research
Medical knowledge is expanding at an unprecedented rate, with thousands of new studies published daily. Systematic reviews, which are comprehensive assessments that analyse all available evidence on a specific medical question, are essential for making informed healthcare decisions. However, these reviews traditionally take about two years to complete and cost approximately $350,000 each.
"By the time many systematic reviews are published, they're already outdated," explains Professor Zuccon.
"This creates a fundamental problem for evidence-based medicine, where decisions should be based on the most current research," highlights Associate Professor Koopman.
With AI-driven solutions, we can transform how systematic reviews are conducted. “The advancement of AI is a game-changer for overcoming these challenges,” says Dr. Leelanupab.
How AI Changes the Game
The new platform combines three cutting-edge technologies to dramatically accelerate the review process:
Large Language Models (LLMs): Similar to the technology behind ChatGPT, these AI systems can understand and analyse text from medical studies, helping to identify relevant information quickly.
Dense Retrieval Systems: Unlike traditional keyword searches, these advanced search engines understand the meaning and context of medical studies, helping to find relevant research that might otherwise be missed.
Active Learning: The system gets smarter with use, learning from reviewer feedback to continuously improve its accuracy.
Practical Applications
The platform offers two main components:
- A user-friendly web tool that allows medical professionals and researchers without technical expertise to conduct effective and efficient systematic reviews with the support of AI technologies
- An open-source programming library that enables researchers to customise and develop new models, fostering open research.
Dr Leelanupab says the screening phase is particularly time-consuming, requiring researchers to review thousands of studies manually.
“This technology could reduce months of work to just days or weeks,” he says.
Benefits Beyond Speed
While faster reviews are the most apparent benefit, the platform offers additional advantages:
- Improved reproducibility: The systematic approach ensures consistent results
- Greater accessibility: Organisations with limited resources can now conduct high-quality reviews
- Cost reduction: The estimated $350,000 per review could be significantly decreased
- Better healthcare decisions: More timely evidence leads to better patient outcomes
Looking Ahead
As medical research continues to accelerate, tools like this will become increasingly important. Organisations funding medical research are already taking notice, as faster reviews mean their investments can impact healthcare more immediately.
The platform is designed to complement rather than replace human expertise.
"The goal isn't to automate systematic reviews entirely," Dr Leelanupab says. "It's to handle the repetitive aspects so that medical experts can focus on interpretation and application."
For patients, healthcare providers, and policy makers alike, this technology promises a future where medical decisions are based on more complete and current evidence – potentially improving healthcare outcomes for everyone.
If you're interested in learning more about this groundbreaking research project, you can contact Dr Teerapong Leelanupab on t.leelanupab@uq.edu.au
