Objective: We assessed whether the three-dimensional fluid-attenuated inversion-recovery (3-D FLAIR) and three-dimensional inversion-recovery with real reconstruction (3-D real IR) sequences can be used to detect cochlea endolymphatic hydrops (EHs) in guinea pigs using 3 Tesla magnetic resonance imaging (3T MRI). The results of 3-D real IR imaging were compared with histopathological outcomes.
Materials and methods: Fourteen healthy men and women albino guinea pigs were used in this study. Their right ears received procedures that promoted EHs, and their left ears were used as untreated controls. High-resolution 3T MRI, combined with the intratympanic injection of gadolinium (Gd) in both ears, was performed 8 to 12 weeks after surgery. Both sides of the cochlea midmodiolar sections were observed under a light microscope and saved as the histopathological images. The signal-to-noise ratios (SNRs) and contrast-to-noise ratios (CNRs) between the T2-weighted 3-D FLAIR and T2-weighted 3-D real IR sequences were compared. The appearance of EHs in the basal, second, third, and apical turns of the cochlea was further evaluated using 3-D FLAIR, 3-D real IR, and the histopathological images. Moreover, the maximum scala media area ratios (SMRs) on the histopathological sections were compared with the grading of the EHs on the 3-D real IR sequence with regard to each turn of the cochlea.
Results: Significant differences were found between the 3-D FLAIR and 3-D-real IR sequences with regard to the SNRs and CNRs (p < 0.05): the 3-D FLAIR sequence exhibited higher SNRs (SNRROI: 347.95 ± 105.01; SNRB: 103.28 ± 17.61) compared with the 3-D real IR sequence (SNRROI: 86.71 ± 30.11; SNRB: 11.11 ± 3.45), whereas the 3-D real IR sequence showed higher CNRs (2.78 ± 0.58) compared with the 3-D FLAIR sequence (2.18 ± 0.55). Various degrees of EHs were observed in each turn of the cochlea in the experimental ears on the basis of the histopathological images. Thirteen, 10, 4, and 0 EHs were observed in the basal, second, third, and apical turns of the cochlear using 3-D FLAIR images, respectively, whereas 14, 14, 14, and 13 EHs were found using 3-D real IR images, respectively. Significant differences were found between the two sequences when evaluating the second, third, and apical turns of the cochlear but not with regard to the basal turn (p < 0.05). The SMRs were proportional to the extent of the EHs on 3-D real IR imaging in each turn of the cochlea.
Conclusions: 3-D real IR images are clearer than 3-D FLAIR images, and they can display cochlea EHs more precisely using 3T MRI in guinea pigs. The extent of the EHs on 3-D real IR imaging was more consistent with the histopathological observations in each turn of the cochlea.