Peripheral nerve MRI: precision and reproducibility of T2*-derived measurements at 3.0-T : a feasibility study.

Abstract

To prospectively evaluate the precision and reproducibility of T2*-derived measurements of the peripheral nerves. The study was approved by the local ethics committee and written informed consent was obtained. Bilateral upper and lower limb MRI examination was performed in 40 healthy subjects on a 3.0-T scanner. MRI protocol included T1-turbo spin-echo, T2-turbo spin-echo with fat suppression, and multiecho gradient recalled echo. Measurements of T2* times on T2* maps at different anatomical levels were performed. Three authors measured independently and in different sessions at baseline and after 4 weeks. Non-parametric tests and Bland-Altman statistics were used. Minimum and maximum percentage variability were 10 % and 19 % for T2* (84-91 % of reproducibility). Maximum values of minimum detectable differences between limbs was 16 % (with 95 % CI: 2-37). Intra- and inter-observer agreement of the three radiologists for T2* was considered good. Evaluating the combined influence of the observer and of the repeated measurements the reproducibility was 87-98 %. T2* measurement of the peripheral nerves is precise and reproducible. The healthy contralateral side can be used as an internal control. Variations in T2* values up to 16 % have to be considered.