Capability of identifying red nuclei in different pulse sequences of regular 1.5-tesla magnetic resonance images

To investigate the optimal pulse sequences of commonly used 1.5-tesla MRI for identifying the red nucleus (RN) to aid targeting of the subthalamic nucleus (STN).

Forty-six healthy adults were enrolled for this prospective study. All subjects underwent MR studies of 5 sequences: diffusion-weighted imaging (DWI), T(1)-weighted fluid-attenuated inversion recovery (T1IR), fast spin echo T(2)-weighted imaging (FSE-T2WI), T(2)-weighted fluid-attenuated inversion recovery (T2FLAIR) and T(2)*-weighted gradient-echo (T2*-GRE) sequences. The clearness of the RN contour was analyzed.

Overall, the RN was identified in 98% subjects without gender and age differences. The RN was demarcated on a 5-mm slice relatively better in T2FLAIR (93.5%), followed by FSE-T2WI (78.3%), T2*-GRE (65.2%) and DWI (43.5%) sequences, but was completely invisible on the T1IR image. Generally, the signal intensity in all MR sequences decreased mildly on 2-mm slices with a similar identifying power. The detecting power on 5-mm slices was in favor of T2FLAIR with 94% sensitivity, 10% specificity, and 1.89 odds ratio compared to FSE-T2WI. In addition, the scanning time of T2FLAIR was longer in comparison to the FSE-T2WI study.

T2FLAIR is an alternative to FSE-T2WI that can readily demarcate the RN to help target the STN.