The purpose of this study was to compare the relative usefulness of multish
ot turbo spin echo (TSE) and half-Fourier single-shot turbo spin echo (HAST
E) far determination of optimal breath-hold fast Ta-weighted technique in t
erms of lesion detection, lesion-to-liver contrast-to-noise ratio (CNR), an
d image quality. The images of TSE with and without fat suppression (FS) an
d of HASTE with and without FS were retrospectively reviewed for 49 patient
s with 128 lesions. Without FS, TSE and HASTE images allowed depiction of f
ocal hepatic masses (112 of 128, sensitivity = 87.5%) at the same rate. TSE
with FS depicted more focal lesions (115 of 128, 89.8%) than HASTE with FS
(109 of 128, 85.2%), but the difference was not statistically significant
(P >.05), The CNR of each lesion on HASTE sequences was greater (P <.01) th
an that on TSE sequences. The CNR of hemangioma was distinct from that of s
olid tumors and cystic lesions in all sequences, and the range of CNR in ea
ch group of pathologies overlapped less and were well separated in the HAST
E sequences. HASTE sequences produced better image quality with fewer artif
acts (P <.0001), The results of this study suggest that HASTE sequences all
ow differentiation between solid tumors, hemangiomas, and cystic lesions in
terms of CNR, producing fewer image artifacts, with acceptable sensitivity
in lesion detection.