Relative intensity of abdominal organs in MR images

Richard L. Ehman, Bent O. Kjos, Hedvig Hricak, Robert C. Brasch, Charles B. Higgins

Research output: Contribution to journalArticlepeer-review

40 Scopus citations


Knowledge of the normal relative intensity of organs and tissues is a valuable aid to clinical interpretation of magnetic resonance images. In this study the in vivo spin echo image intensities of normal parenchymal organs and other structures in the upper abdomen were evaluated for eight parameter combinations. The examinations of 40 patients were used. Image intensity and calculated Tl, T2, and spin density values were obtained for liver, spleen, pancreas, renal cortex, renal medulla, bone marrow, skeletal muscle, and fat. Repetition times (TR) of 500, 1,000, 1,500, and 2,000 ms and echo times of 28 and 56 ms were used. The Tl and T2 values and relative spin density were calculated using a new algorithm. Liver had the smallest relative standard deviation of Tl of all the tissues studied. For comparison purposes, relative image intensities were calculated by normalizing them to the intensity of liver in the same image. The resulting compiled data show the normal ranks and ranges for relative intensity for the tissues in each of eight types of spin echo images. Although images with short TR and echo time (TE) are known to display the greatest Tl contrast, the mean relative intensities of all tissues except muscle and fat in the TR = 500 and TE = 28 ms images were within 20% of liver. A much larger spread in the normal relative intensities was observed with longer TE and TR.

Original languageEnglish (US)
Pages (from-to)315-319
Number of pages5
JournalJournal of computer assisted tomography
Issue number2
StatePublished - 1985


  • Abdomen
  • Nuclear magnetic resonance
  • Nuclear magnetic resonance
  • Techniques

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging


Dive into the research topics of 'Relative intensity of abdominal organs in MR images'. Together they form a unique fingerprint.

Cite this