The STIR sequence in MR imaging.

A recent paper on MR imaging of infectious spondylitis [1) elaborated on and substantiated a previous report [2) of STIR in the evaluation of spinal infection. However, it also contained an important error in pulse sequence nomenclature: STIR was described as "short T1 inversion recovery. " The STIR sequence was described and named by Drs. Bydder, Young, and Steiner [3 , 4) as "short Tl inversion recovery." Typically this technique uses Tl (time to inversion) values of 100-160 msec vs the earlier convention of long Tl times of 400-600 msec-hence its name. In the article by Thrush and Enzmann [1), STIR is repeatedly referred to as "short T1 inversion recovery." We use STIR extensively in our practice [2, 5) and frequently have had to correct well-intentioned copy editors who change "TI " to "T1 " because of their lack of familiarity with inversion-recoyery imaging sequences. Many such errors undoubtedly have entered the literature because of oversights in proofreading. However, I have seen this error with increasing frequency in a variety of publications, including peer-reviewed journals. As STIR is used for detection of substances with long T1 and long T2 [3). this misnomer is not only inaccurate but misleading. STIR is also occasionally referred to as the "short tau inversion recovery" sequence. This is also incorrect. STIR is an inversion-recovery-spin-echo sequence with a TE of 30-40 msec; therefore the tau time (TE/2) is in fact rather long, not short. I am concerned that this incorrect nomenclature may be accidentally perpetuated by authors, reviewers, editors, and readers who have an incomplete understanding of the STIR sequence. I encourage all who refer to STIR in their publications to carefully review the original articles [3, 4) and to refer to this sequence by its correct and descriptive name. Careful proofreading of articles is also necessary to ensure that such errors do not inadvertently become accepted into the scientific literature.