Abstract
We evaluated the feasibility of long ETL, 3D FSE sequences, maximum intensity pixel (MIP) projection and surface rendering reconstructions in high resolution visualisation and diagnosis of normal and abnormal inner ear anatomy, in acceptable acquisition times. Five healthy children and nine children with congenital inner ear malformations were imaged using a 1.5-T MR scanner and 5 inch phased array receiver coils. Forty 0.7 mm-thick coronal sections obtained with a 3D FSE sequence (TR/TEeff: 4000/150, ETL: 31, 512×256 matrix, 20 cm 3/4 FOV, 1 NEX, scan time: 8:30 minutes) were reformatted (0.43 mm thickness, no gap). Anatomical structure visibility was scored. Oblique reformations, MIP projections and 3D surface displays (volume rendering) of all inner ears were generated on a SUN workstation using GE software. T2-weighted 3D FSE images demonstrated four nerves in the internal auditory canal, AICA loop, vestibule, semicircular canals, cochlea, modiolus, osseous spiral lamina, and the cochlear aqueduct were in all normal ears. Facial nerve visibility included labyrinthine segment (92% of cases), tympanic (96%), and mastoid (100%). Vestibular aqueduct was seen in 64%. MIP images always depicted cochlear turns, semicircular canals and ampullae. MIP and stereoscopic 3D segmented images provided a clearer understanding of extent of ear malformations. T2-weighted 3D FSE imaging enables visualisation of submillimeter inner ear structures with acceptable scan times. Volumetric imaging and segmentation analysis of the labyrinth improves the understanding of normal and abnormal inner ear anatomy in vivo, paving the way for 3D planning of intervention and quantitative volumetric studies.
Get full access to this article
View all access options for this article.