Identification and characterization of tissues by T2‐selective whole‐body proton NMR imaging

Publisher Website
Google Scholar
Abstract
T2-selective proton imaging at 10 MHz is based on a Carr-Purcell-Meiboom-Gill pulse sequence with at least 36 echoes and 2τ ≃ 6 ms employing the projection reconstruction technique. The transverse magnetization decay of the volume elements is multiexponential composed of up to three monoexponential T2 components. The T2 values are arranged in T2 histograms which are slice specific. T2 windows are defined in the T2 histogram and used for imaging and for selecting T2 values which are assigned to proton classes and tissues. Tissue characterization is based on the visualization of differences in T2 behavior. The T2 values are transformed into 256 gray levels or, by use of a red-green-blue lookup table, into a rainbow code. Two display techniques have been developed and implemented in the RWTH Aachen Magnetic Resonance Software System (RAMSES): (i) for nonselective T2 imaging with secondary T2 selection and encoding of molecular classes on the basis of T2 values; (ii) for T2-selective imaging by employing combined primary and secondary T2 selection and encoding particular tissues and variations therein. The tissues can be identified by selecting different classes of molecules, like water, lipids, and proteins, when shifting the primary T2 window on the T2 axis of the T2 histogram.