Evaluation of meniscal tears: Fast spin-echo versus conventional spin-echo magnetic resonance imaging

Abstract

We compared the performance of fast spin-echo (FSE) with conventional spin-echo (CSE) magnetic resonance (MR) imaging sequences in the detection of meniscal tears. Seventy-three patients underwent MR examination of the knee for suspected internal derangement. Each patient was scanned with a CSE sequence and one of two FSE sequences. The primary difference between the two FSE sequences consisted of the echo train length (ETL). Thirty-seven patients (group 1) were scanned with the FSE I sequence (ETL = 8), and 36 patients (group 2) were scanned with the FSE II sequence (ETL = 4). Menisci were graded as torn or not torn on the basis of their MR appearance. The sequences were compared with each other and with the surgical findings in 31 patients who underwent arthroscopy. In group 1, there was a significant discrepancy between the CSE and FSE I techniques (p = .006). The FSE I sequence detected only 11 of 19 surgically proven torn menisci as opposed to 18 of 19 detected with the CSE sequence. The FSE II sequence performed significantly better in group 2, with an accuracy equal to that of the CSE sequence. FSE sequences are extremely technique dependent with regard to detecting meniscal tears and should not replace CSE sequences in this setting until further studies are performed to optimize the technique.