MRI shown doable in a third of the time with no compromise on quality

Fifteen minutes can seem like an eternity to some patients undergoing MRI exams. Researchers have shown that some scans can be done in about a third of that time, and with no significant falloff in diagnostic performance. Their findings were published online Jan. 31 in the American Journal of Roentgenology.

Naveen Subhas, MD, of the Cleveland Clinic and colleagues conducted a retrospective review of 151 shoulder MRI exams on a 3-T scanner and 50 shoulder MRIs on a 1.5-T scanner.

They compared the interpretations of three musculoskeletal radiologists from two academic medical centers reading MR images acquired in four fast (around five minutes) and five standard (around 15 minutes) scanning sequences, measuring sensitivity and specificity for 51 patients who underwent surgery.

The primary indications for the MRI were for pain not otherwise specified (n = 66), rotator cuff tear (n = 49), labral tear or biceps tendon abnormality (n = 22), instability or dislocation (n = 8) and other injuries and conditions.

The team reports that their experimental fast shoulder MRI protocol was much shorter than the standard MRI protocol, clocking in at an average time of five minutes 23 seconds compared with 14 minutes six seconds (62 percent reduction) at 3-T and four minutes 30 seconds compared with 15 minutes 39 seconds (71 percent reduction) at 1.5-T.

As for performance, they found the fast MRI protocol to be interchangeable with standard MRI with no significant decrease (less than 5 percent) in the rate of interreader agreement when standard MRI was replaced with fast MRI.

Additionally, Subhas et al. report, fast and standard MRI had similar interreader kappa agreement for all evaluated structures, and readers’ ability to identify major findings on fast and standard MRI was not significantly different.

Further, the sensitivity and specificity of fast MRI in diagnosing rotator cuff and biceps tendon tears, labral tears and cartilage defects were nearly identical to the accuracy of standard MRI, the authors write.

Fast five-minute shoulder MRI with standard 2D fast spin echo sequences using parallel imaging “is interchangeable, with similar interreader agreement and accuracy, with standard shoulder MRI for evaluating shoulder injuries,” they conclude.

Among the limitations in study design the authors acknowledge is the relatively older age of the patients whose scans were studied. Those in the 3-T group had a mean age of 46.95 years, while those in the 1.5-T group had a mean age 53.74 years.

“[T]he results may not be generalizable to a younger patient population, among whom the frequency of labral abnormalities is higher and the frequency of rotator cuff disease is lower,” the authors write.

That said, the high diagnostic performance of the team’s fast shoulder MRI protocol may point toward a better MRI experience for all involved.  

“Routine use of this protocol could reduce the current 30- to 40-minute time slot for shoulder MRI (20–30 minutes for imaging and 10 minutes for setup) to perhaps a 15-minute time slot (five minutes for imaging and 10 minutes for setup), effectively doubling the number of shoulder MRI scans that could be performed in a day for better utilization of the scanner,” the authors write.

The reduction in exam time “may also improve the tolerability and comfort of the examination, especially when imaging anxious or claustrophobic patients,” they add. “Finally, the fast shoulder protocol could be used as a tool to reduce patient wait times when there is a backlog of cases.”