The aim of the study was to assess whether a 3-D FSE sequence with rea
l-time multiplanar reconstruction could replace our standard 2-D imagi
ng of the cervical spine, MRI was performed on a GE Vectra 0.5 T syste
m using a quadrature surface neck coil, Signal intensity of CSF, spina
l cord, surrounding tissue and background were measured, Contrast to n
oise ratio and contrast between the different tissues was calculated f
or the different sequences, A subjective assessment of the various les
ions was made, No statistical difference in tissue contrast was found
between 2-D and 3-D images when the contrast between cord and CSF, or
between cord and marrow was calculated. Contrast between cord and marr
ow was better on 3-D images, The contrast to noise ratio was better on
2-D images compared with the 3-D images for both cord/CSF and CSF/mar
row but there was no difference between the 2-D and 3-D images for cor
d/marrow, In three patients the lesion was better demonstrated on the
3-D sequence than on the 2-D combination, but only in one of these was
the abnormality not visible on the 2-D images; in six patients the 2-
D images were considered superior, Although the 3-D sequence reduced o
verall imaging: time we found that at 0.5 T contrast was inadequate, a
nd that lesions were less clearly demonstrated than on the 2-D sequenc
es, We conclude that a single 3-D sequence cannot satisfactorily repla
ce the 2-D combination routinely used at 0.5 T.