Better diagnoses of back pain may be possible with a new magnetic resonance imaging (MRI) scanning techniquefor analyzing degenerated spinal discs. Improving accuracy from 70% using current methods to 97%, the Decay Variance technique could be in clinical use within four years, say researchers who developed the technique at the University of New South Wales in Sydney, Australia.
Estimating intervertebral disc (IVD) degeneration that contributes to chronic lower back pain is currently challenging using MRI. Novel postprocessing algorithms for MRI data represent an option for performing nondestructive and objective assessments of IVD decay and regeneration.
Researchers in the Spine Service department of St. George and Sutherland Clinical School confirmed that the variability in the rate of signal decay over a multi-echo MRI sequence can help distinguish between a healthy and a degenerated IVD.
In a laboratory study imaging 25 rabbits, published online August 1, 2019 in JOR Spine, the researchers, led by Kyle Sheldrick, MD, confirmed that the decay variance maps created with a multi-echo T2 MRI sequence quantitatively assessed disc degeneration more accurately and with less image-processing time than quantitative T2 relaxometry.
Images generated by the Decay Variance technique showed well-defined dark intervertebral disc regions against a mid-gray vertebral body, and well-demarcated dark nucleus pulposus (NP) regions against a bright white annulus firosus (AF). The central dark NP became increasingly homogeneous with the AF with increasing degeneration. In addition to taking only one second to produce, the Decay Variance maps had superior sensitivity and specificity for the detection of degenerate discs, compared to T2 signal intensity or quantitative T2 mapping.
Comparing the new technique with that currently used by radiologists, lead author Ashish Diwan, MD, PhD, director of spine service at the Department of Orthopaedic Surgery, says the difference represents a shift in the way the data is analyzed.
“Current techniques see how quickly or slowly nuclei lose alignment after a strong burst of radio waves,” he says. “Instead of trying to figure out how fast or slow this signal decay is, our technique measures whether the nuclei in a sample are lining up at the same speed as each other over the course of a scan, or a range of different speeds – hence decay variance.”
“In research, MRI scans of the spinal discs to diagnose degeneration are assessed quantitatively by applying computer algorithms to determine the ‘true’ value in each pixel and adjust for random measurement error/noise, because this has been shown to be more accurate than using raw values,” Dr. Diwan said in an interview with Applied Radiology.
“However, in general clinical use, the simple measured raw value is used because existing quantitative algorithms take a very large amount of time to process. Our team has been working to develop faster processing algorithms that are at least as accurate and not tied to any particular MRI vendor,” he said. “In a research MRI scanner in rabbits, our algorithm is between 5 and 1,800 times faster than existing postprocessing algorithms. It is also moderately more accurate in the diagnosis of disc degeneration in contrast to most commercial proprietary algorithms.”
The team plans an observational case control study to test whether the same computation postprocessing speed is observed and acceptable accuracy is achieved on a conventional MRI scanner in patients with and without back pain, says Dr. Diwan.
“Our hypothesis is that we will be able to process the same images that existing techniques use, but in no more than one-fifth of the time, while still maintaining the same accuracy as existing quantitative algorithms. This is important because if we can reduce the time taken to perform computer postprocessing of whole lumbar spine images from 30 minutes to 5 minutes or less, it will remove the main barrier to quantitative MRI actually entering clinical practice,” he says.
“Repeated surveys of radiologists show that the main barrier to introducing better techniques to practice is that the amount of time taken to process the scans is so long that it is clinically impractical,” he adds.
Potential new MRI technique to assess lower back pain. Appl Radiol.