Three dimensional MR gradient recalled echo imaging of the inner ear: comparison of FID and echo imaging techniques

The detailed structures of the inner ear make this region a diagnostic challenge for radiologists. Thin section high resolution CT is the "gold standard" for studies of the fine bony detail of the inner ear. Although CT can delineate bony structures, fine soft tissue details surrounded by CSF/endolymph (such as nerves in the internal auditory canal) are not easily identified. Conventional MR spin-echo T2-weighted images provide good image contrast for such structures, but the current commercially available minimum slice thickness of approximately 2-3 mm is too thick for the inner ear. Volume gradient recalled echo (GRE) MR imaging techniques can be used to achieve thin slices (< 2 mm) while maintaining adequate contrast for detailed examination. In the work reported here a volume GRE sequence that images the echo formed in a steady-state-free-precession (termed "CE-FAST" or "SSFP" on various commercial MRI systems and called SSFP-echo in this work) was used to image inner ear structures. This technique was compared with images generated using conventional volume GRE techniques (GRASS). While small flip angle volume GRE imaging has been used for inner ear imaging previously, the low contrast typical of such density weighting makes it difficult to distinguish soft tissue structures from surrounding CSF/endolymph. In this work, contrast-to-noise ratios (CNR) between CSF/endolymph and brain parenchyma were compared between the sequences at 15 degrees, 30 degrees, 60 degrees, and 90 degrees flip angles. The SSFP-echo sequence produced higher CNR for such structures and consistently outperformed GRASS sequences at flip angles of 30 degrees, 60 degrees, and 90 degrees.(ABSTRACT TRUNCATED AT 250 WORDS)