Motion-related repeat MRI sequences create considerable operational burden
Motion-related repeat MRI sequences can create considerable operational burden for radiology providers, according to new research.
Despite the modality’s clinical utility, MRI’s long acquisition times can make it vulnerable to patient movement, with such motion artifacts a common cause of poor image quality. In regular practice, the typical response is to repeat the scan—which can, in turn, reduce throughput and contribute to downstream delays, experts write in JACR.
Previous studies have estimated that upward of $115,00 in per-scanner revenue is lost each year, due to motion-driven MRI repeats. Researchers with the NYU Grossman School of Medicine and Siemens Healthineers aimed to better understand the impact of these do-overs at the New York institution.
They uncovered measurable operational burden, with diagnostic yield from repeats varying substantially, depending on the anatomic region and sequence type.
“These findings support a more targeted approach to motion mitigation in which health systems track not only repeat frequency and scanner time, but also the diagnostic yield of repeat acquisition, enabling sequence-specific strategies to reduce low-value repeat imaging and preserve scanner capacity,” corresponding author Hersh Chandarana, MD, with the Department of Radiology at NYU Grossman, and colleagues wrote Sept. 22.
The retrospective study analyzed six months of MRI exams on scanners with sequence-level analytics. Exams with technologist-labeled, repeated sequences were identified in NYU’s records. Two readers scored scans for image quality and diagnostic confidence, indicating whether another repeat would be requested.
Among over 85,000 MRI exams during the study period, about 5% (or 4,000) included at least one repeated sequence. Re-scans were highest among pediatric patients (10%) and lowest in breast MRI (4%). In a closer analysis of brain MRIs, 87 baseline-repeat pairs from 77 patients were evaluated. Repeat acquisition significantly improved image quality and diagnostic confidence across all metrics. However, about 20% of repeat sequences did not improve; these gains declined with greater lapses in exam time, researchers found.
Another analysis of magnetic resonance cholangiopancreatography scans included 30 baseline-repeat pairs from 27 patients. Repeat acquisitions did not significantly improve image quality nor diagnostic confidence. Both readers in the study would have requested another repeat for about half of these do-over acquisitions. Motion-driven repeat MRI sequences accounted for about 115 additional scanner hours, corresponding to about 268 lost MRI slot-equivalents over six months.
“These findings have practical implications for MRI operations,” the authors charged. “Repeat rate alone is an incomplete quality metric, because repeat acquisition is not a uniformly successful corrective action. Sequence-level analytics can help track not only which protocols generate repeats, but also which repeat acquisitions improve diagnostic quality, remain degraded, or require additional acquisition. Such tracking could support actionable interventions, including re-ordering protocols to acquire diagnostically critical or motion-sensitive sequences earlier in the examination, prioritizing sequences with low repeat success rates for motion-robust acquisition or AI-based motion correction, and developing decision-support tools that identify when an additional repeat is unlikely to improve image quality and rescheduling, or alternative imaging may be more appropriate.”
Read more, including potential study limitations, in the Journal of the American College of Radiology.
Marty Stempniak has covered healthcare since 2012, with his byline appearing in the American Hospital Association's member magazine, Modern Healthcare and McKnight's. Prior to that, he wrote about village government and local business for his hometown newspaper in Oak Park, Illinois. He won a Peter Lisagor and Gold EXCEL awards in 2017 for his coverage of the opioid epidemic.