Simultaneous T2 and lipid quantitation using IDEAL-CPMG

Robert L. Janiczek, Giulio Gambarota, Christopher D.J. Sinclair, Tarek A. Yousry, John S. Thornton, Xavier Golay, Rexford D. Newbould

Research output: Contribution to journalArticlepeer-review

37 Scopus citations

Abstract

Muscle damage, edema, and fat infiltration are hallmarks of a range of neuromuscular diseases. The T2 of water, T2,w, in muscle lengthens with both myocellular damage and inflammation and is typically measured using multiple spin-echo or Carr-Purcell-Meiboom-Gill acquisitions. However, microscopic fat infiltration in neuromuscular diseases prevents accurate T2,w quantitation as the longer T2 of fat, T 2,f, masks underlying changes in the water component. Fat saturation can be inconsistent across the imaging volume and removes valuable physiological fat information. A new method is presented that combines iterative decomposition of water and fat with echo asymmetry and least squares estimation with a Carr-Purcell-Meiboom-Gill-sequence. The sequence results in water and fat separated images at each echo time for use in T2,w and T 2,f quantification. With knowledge of the T2,w and T 2,f, a T2-corrected fat fraction map can also be calculated. Monte-Carlo simulations and measurements in phantoms, volunteers, and a patient with inclusion body myositis are demonstrated. In healthy volunteers, uniform T2,w and T2-corrected fat fraction maps are present within all muscle groups. However, muscle-specific patterns of fat infiltration and edema are evident in inclusion body myositis, which demonstrates the power of separating and quantifying the fat and water components.

Original languageEnglish (US)
Pages (from-to)1293-1302
Number of pages10
JournalMagnetic resonance in medicine
Volume66
Issue number5
DOIs
StatePublished - Nov 2011

Keywords

  • edema
  • muscular dystrophy
  • water-fat separation

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Fingerprint

Dive into the research topics of 'Simultaneous T2 and lipid quantitation using IDEAL-CPMG'. Together they form a unique fingerprint.

Cite this