Compressed sensing reduces ankle MRI time by 20%

Compressed sensing (CS) can reduce the typically long acquisition time of conventional ankle MRIs while preserving imaging quality, according to a small study published in the European Journal of Radiology.

Due to the inherent complexity of the ankle, Two-dimensional (2D) turbo spin echo (TSE) MR sequences must be acquired in all three dimensions when imaging the joint. However, the standard imaging protocol comes with long exam times and limited resolution, according to Alexandra S. Gersing, MD, from Technical University of Munich’s Department of Radiology in Germany, and colleagues.

Therefore, the team sought to determine if CS—"a technique that acquires less data through k-space random undersampling and enables a reduction in total acquisition time…”—could reduce overall scan time while maintaining diagnostic image quality.

The researchers scanned 20 patients with a mean age of 30.2 years old at 3T MRI. Images were acquired using parallel imaging based on sensitivity encoding (SENSE) alone and when using a combined method of CS and SENSE. Two radiologists graded image quality on a 5-piont Likert scale and assessed the signal-to-noise (SNR) ratio and contrast-to-noise (CNR) ratio of different anatomical structures in the ankle scans.

Overall, CS reduced the acquisition time of anatomical 2D turbo spin echo MR imaging by 20%. The radiologists found “substantial to perfect” agreement between images produced with SENSE only and the combination of CS and SENSE when assessing cartilage, subchondral bone and ligaments, the authors concluded.

SNR was barely higher in the combined method compared to SENSE alone, but, according to the researchers, the finding was not significant. Similarly, CNR of the cartilage/fluid, subchondral bone/cartilage, ligaments/fluid and ligaments/fat did not differ significantly between CS and SENSE and SENSE alone. Interreader agreement was, again, “substantial to excellent” for both techniques.

“Moreover, there was no image blurring detected as in previous studies assessing 3D joint imaging with compressed sensing,” Gersing el. added.

There were a few limitations of this study, including a lack of arthroscopy as a standard of reference and the fact that few participants showed pathologies at their ankle.

“Additional studies with following arthroscopy as a standard of reference are needed in order to assess the diagnostic performance of these high-resolution sequences acquired with compressed sensing at the ankles of symptomatic patients,” the researchers explained.