Cavernous hemangioma of the liver: Assessment of MR tissue specificity with a simplified T2 index

Scott A. Mirowitz, Joseph K.T. Lee, Jay P. Heiken

Research output: Contribution to journalArticlepeer-review

36 Scopus citations

Abstract

This is a comparison of the ability of two quantitative magnetic resonance (MR) indices: (a) the second echo signal drop (SESD) (a simplified index of T2 relaxation times that we have developed); and (b) the previously described lesion/liver signal intensity ratio (LLR) to provide histologic diagnosis in hepatic lesions >2 cm in diameter. In 55 patients 108 hepatic masses [31 cavernous hemangiomas (CH), 68 metastases, and 9 hepatocellular carcinomas] were scanned at 0.5 T. Statistically significant differences between CH and malignant lesions were obtained in mean SESD (p=0.0006) and LLR (p = 0.0008) using repetition time (TR) 2,100/echo time (TE) 35, 60 ms. Application of cutoff values derived from receiver-operator characteristic analysis led to a correct diagnosis in 100 and 94% of lesions, respectively. Using TR 2,100/TE 35, 90 ms in a different patient population, CH and malignancies again displayed significantly different mean values, using the SESD (p = 0.0090) and LLR (p = 0.0024) methods. These measurements provided a correct diagnosis in 74 and 81%, respectively. Accuracy was increased in those cases in which the diagnosis by SESD and LLR were concordant. When compared with visual analysis, these quantitative methods appear to achieve near 100% accuracy in the differentiation of hepatic CH from malignancies.

Original languageEnglish (US)
Pages (from-to)223-228
Number of pages6
JournalJournal of computer assisted tomography
Volume14
Issue number2
DOIs
StatePublished - Mar 1990

Keywords

  • Liver
  • Liver, Neoplasms
  • Magnetic resonance imaging
  • Techniques

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Fingerprint

Dive into the research topics of 'Cavernous hemangioma of the liver: Assessment of MR tissue specificity with a simplified T2 index'. Together they form a unique fingerprint.

Cite this