Evaluation of the menisci of the knee joint using three-dimensional isotropic resolution fast spin-echo imaging: diagnostic performance in 250 patients with surgical correlation.

Abstract

To compare the diagnostic performance of FSE-Cube, a three-dimensional isotropic resolution intermediate-weighted fast spin-echo sequence, with a routine magnetic resonance (MR) protocol at 3.0 T for detecting surgically confirmed meniscal tears of the knee joint in a large patient population. FSE-Cube was added to a routine MR protocol performed at 3.0 T on 250 patients who underwent subsequent knee arthroscopy. Three radiologists independently used FSE-Cube during one review and the routine MR protocol during a second review to detect medial and lateral meniscal tears. Using arthroscopy as the reference standard, the sensitivity and specificity of FSE-Cube and the routine MR protocol for detecting meniscal tears were determined for all readers combined. McNemar's tests were used to compare diagnostic performance between FSE-Cube and the routine MR protocol. FSE-Cube and the routine MR protocol had similar sensitivity (95.5%/95.3% respectively, P = 0.94) and similar specificity (69.8%/74.0% respectively, P = 0.10) for detecting 156 medial meniscal tears. FSE-Cube had significantly lower sensitivity than the routine MR protocol (79.4%/85.0% respectively, P < 0.05) but similar specificity (83.9%/82.2% respectively, P = 0.37) for detecting 89 lateral mensical tears. For lateral meniscal tears, FSE-Cube had significantly lower sensitivity (P < 0.05) than the routine MR protocol for detecting 19 root tears but similar sensitivity (P = 0.17-1.00) for detecting all other tear locations and types. FSE-Cube had diagnostic performance similar to a routine MR protocol for detecting meniscal tears except for a significantly lower sensitivity for detecting lateral meniscal tears, which was mainly attributed to decreased ability to identify lateral meniscus root tears.