ARTIFICIAL INTELLIGENCE DETERMINING WHY SOME PEOPLE WITH WHIPLASH SUFFER LONG-TERM
In a technological breakthrough, researchers are using artificial intelligence to determine why some people with whiplash will go on to have long-term changes in their muscles. A team of researchers across the world, led by Kolling Institute’s Professor James Elliott, Dr. Andrew Smith (Regis University) and Dr. Ken Weber (Stanford University), have piloted a new method to quickly and accurately analyse complex muscles traversing the cervical spine to identify those patients at risk of a slower recovery following a whiplash injury. In the process, they have demonstrated a way to reduce the time it currently takes to analyse the imaging from hours to just seconds, opening the way for the technology to be used widely in radiology research and clinical practice. Researchers have used advanced Magnetic Resonance Imaging (MRI) to quantify specific structural changes – or fat infiltration -- within the muscles of the neck with the aim of identifying those who may go on to suffer long-term whiplash injuries. But Prof Elliott, of The University of Sydney, said this method was not without challenges as it requires patients to undergo an expensive MRI and not everyone needs an MRI. The measure further requires an individual to ‘draw around muscles’; a process known as manual segmentation. “Drawing circles around the neck muscles of interest, could take the most- experienced researcher up to two hours to complete and that is on a good day,” Prof Elliott said. “There is a need to enhance, if not automate, the segmentation
process, and computers (or deep learning artificial intelligence) is a great way to do so.” Deep learning is a supervised machine learning method in the field of artificial intelligence used to solve complex pattern recognition problems. It is the same technology used by a smart phone to recognise a face or a voice. Supervised deep learning algorithms can help analyse and interpret medical images by preserving spatial information and greatly reducing the complexity of manually drawing containing complex anatomy such as MRIs from the neck can be analysed accurately in seconds rather than hours. “We’ve shown this measure is reliable, clinically friendly and incredible fast,” Prof Elliott said. “It is the first of its kind in the world to provide objective automatic segmentation of muscle markers for the chronic whiplash condition. “It’s now a matter of expanding this available technology to be used in other common, yet enigmatic musculoskeletal conditions such as, low back circles on each scan. This means that scans
pain or rotator cuff injuries
Professor Jim Eilliott who is a world expert on whiplash injuries
involving the shoulder.” “Unfortunately, there is no gold- standard diagnostic test to identify such patients as imaging tests have not consistently revealed the biological cause(s) for the wide and varied symptoms,” he said. “We have been able to use this new methodology to look at unique patterns identified in the muscles and this may lend itself to the identification of which patients are most susceptible to problems in the long-term.”
Professor Jim Eilliott conducting an MRI
WWW.KOLLINGINSTITUTE.ORG.AU 5
Made with FlippingBook - professional solution for displaying marketing and sales documents online