RT Journal Article
SR Electronic
T1 Brain Abscess and Necrotic Brain Tumor: Discrimination with Proton MR Spectroscopy and Diffusion-Weighted Imaging
JF American Journal of Neuroradiology
JO Am. J. Neuroradiol.
FD American Society of Neuroradiology
SP 1369
OP 1377
VO 23
IS 8
A1 Ping H. Lai
A1 Jih T. Ho
A1 Wei L. Chen
A1 Shu S. Hsu
A1 Jyh S. Wang
A1 Huay B. Pan
A1 Chien F. Yang
YR 2002
UL http://www.ajnr.org/content/23/8/1369.abstract
AB BACKGROUND AND PURPOSE: Discriminating pyogenic brain abscesses from cystic or necrotic tumors is sometimes difficult with CT or MR imaging. We compared findings of proton MR spectroscopy (1H-MRS) with those of diffusion-weighted imaging to determine which technique was more effective for this differential diagnosis.METHODS: Fourteen patients (necrotic or cystic tumor [n = 7]; pyogenic abscess [n = 7]) who underwent 1.5-T 1H-MRS and diffusion-weighted imaging and had findings of ring-shaped enhancement after contrast agent administration were enrolled in this study. Diffusion-weighted imaging was performed with a single-shot spin-echo echo-planar pulse sequence (b = 1000 s/mm2). The apparent diffusion coefficient and ratio were also measured.RESULTS: Spectra for two patients were unacceptable because of either poor shimming conditions or contamination from neighboring fat. Spectra in three of five patients with abscess had lactate, amino acids (including valine, alanine, and leucine), and acetate peaks; one of the three spectra had an additional peak of succinate. In one patient with abscess treated by antibiotics, only lactate and lipid peaks were detected. Spectra for four of seven patients with cystic or necrotic tumors showed only lactate peaks. Lactate and lipids were found in three patients with tumors. Hyperintensity was seen in all the pyogenic abscess cavities and hypointensity in all the cystic and necrotic tumors on diffusion-weighted images.CONCLUSION: 1H-MRS and diffusion-weighted imaging are useful for differentiating brain abscess from brain tumor, but the latter requires less time and is more accurate than is 1H-MRS. 1H-MRS is probably more limited in cases of smaller peripheral lesions, skull base lesions, and treated abscesses.