Morphological characteristics for the magnetic resonance imaging assessment of discoid lateral meniscal tears in children / 中国组织工程研究

ABSTRACT

BACKGROUND: Currently, the morphologic characteristics of the torn discoid lateral meniscus remain unclear, andmorphological indexes used to evaluate the discoid lateral meniscus tears still need to be explored. OBJECTIVE: To discuss the specificity and sensitivity of discoid lateral meniscal tears in children using morphological indexes. METHODS: Seventy-three patients (seventy-six knees) with torn discoid lateral meniscus underwentarthroscopy and MRI data of knee examinations were retrospectively reviewed. All MR images were double-blinded, independently, retrospectively analyzed by attending radiologist (doctor A) and sports physician (doctor B). The following morphological signs of discoid lateral meniscus, such as a internal displaced free edge sign, a hypertrophic free edge sign and a hypertrophied-horn sign, were selected to evaluate the discoid lateral meniscus tears. The likelihood of the discoid lateral meniscus tears was analyzed on basis of arthroscopic findings constituted the gold standard. Sensitivity, specificity and accuracy of MRI diagnosis of the discoid lateral meniscus tears by two physicians were calculated, and the consistency of diagnosis results by two physicians was evaluated using the Kappa statistics.RESULTS AND CONCLUSION: (1) The sensitivity, specificity and accuracy in the diagnosis of discoid lateral meniscustears by internal displaced free edge sign for doctor A were 19%, 94% and 53%, respectively, and for doctor B were 21%, 97% and 55%. (2) The sensitivity, specificity and accuracy in the diagnosis of discoid lateral meniscus tears by hypertrophic free edge sign for doctor A were 52%, 100% and 74%, respectively, and for doctor B were 57%, 97% and 75%. (3) The sensitivity, specificity and accuracy in the diagnosis of discoid lateral meniscus tears by hypertrophied-horn sign for doctor A were 26%, 97% and 58%, respectively, and for doctor B were 24%, 100% and 58%. (4) The sensitivity, specificity and accuracy in the diagnosis of discoid lateral meniscus tears by binding morphological index for doctor A were 86%, 91% and 87%, respectively, and for doctor B were 88%, 94% and 91%. (5) The Kappa statistics for discoid lateral meniscus tears respectively demonstrated a good inter-observer agreement (K > 0.75, P < 0.001). (6) These findings suggest that single morphological index used to evaluate the discoid lateral meniscus tears shows a good specificity, but which also shows a low sensitivity, while a binding morphological index can significantly improve the diagnostic sensitivity. Morphological characteristics may be reliable indexes to evaluate the discoid lateral meniscal tears in children.