NUH develops AI tool to accelerate lumbar spinal stenosis diagnosis
The tech has been trialled in over 50 patients since June.
The National University Hospital (NUH) has developed Spine AI, an artificial intelligence (AI) tool to reduce the time required for radiologists to interpret MRI scans of lumbar spinal stenosis.
The deep learning tool has been trialled in over 50 patients since June, detecting the area where the narrowing of the spinal canal occurs and categorising the severity of stenosis.
Previously, radiologists had to manually evaluate each of the five spinal segments and their five potential stenosis sites, requiring analysis of up to 25 regions per patient.
According to Dr Andrew Makmur, National University Health System’s (NUHS) group chief technology officer, creating detailed reports on stenosis can be labour-intensive and repetitive.
“With Singapore’s ageing population and an expected increase in imaging volume, there is great potential for Spine AI to augment radiologists’ efficiency and allow them to focus on more complex cases,” Makmur said.
Meanwhile, the AI tool was designed in collaboration with a team from the National University of Singapore’s School of Computing (NUS Computing) and the National University Spine Institute.