Abstract

Many of the natural and artificial systems analyzed by 1H-NMR/MRI present populations of nuclei contributing to the signals with distributions of T1 and T2 relaxation times. When working with such systems it can be useful to quantify and/or suppress relaxation time components. Examples are the suppression of fat to improve MRI contrast in tissues, the evaluation of the fat-to-lean ratio in the meat, or the separation of signals from water and oil in porous rocks. A common practice to suppress components is to exploit the differences of T1 (the standard sequences following this approach are STIR, MIR and DIR).

The method called Parametrically Enabled Relaxation Filters with Double and Multiple Inversion (PERFIDI) [1-3] is based on a different line of reasoning: its effect is not to exactly zero the signals for a discrete number of T1 values, but to strongly attenuate the signal in selected ranges of T1 values, while the remaining signals are affected by a computable attenuation, in analogy to the electronic band-pass, high-pass or low-pass filters, emphasizing primarily the components which pass through, rather than those which are blocked. PERFIDI filters are used to separate the sample components according to their T1 values prior to data acquisition rather than doing it mathematically a-posteriori and is particularly useful when a continuous distribution of T1 values is expected. The PERFIDI protocol consists in the acquisition of signals from a chosen RF pulse sequence combined with two or more PERFIDI preambles, and combining linearly the acquired data sets. Each pulse sequence is thus made of two parts. The first one is the preamble, made of a series of inversion pulses, whose timing is varied depending on the desired filter behavior. The second one is the conventional rf sequence to which the filter is applied.

In this presentation we investigate the performance of a PERFIDI protocol based on two-sequences with two-inversion pulses. It is particularly useful in the presence of samples with bimodal distributions of longitudinal relaxation times. Based on the PERFIDI method, we have developed a protocol for very fast determination of the ratio between fat signal and total signal, avoiding the time consuming Inversion-Recovery data acquisition, and applied it to biological samples containing fat. The same method was implemented on ARTOSCAN tomograph (Esaote S.p.A., Genoa, Italy), providing well T1-contrasted images.

References and links:

• Fantazzini P., Sykora S.,
  Italian patent BO2005A000445 registered July 1, 2005 by University of Bologna.
• Sykora S., Bortolotti V., Fantazzini P.,
  PERFIDI: parametrically enabled relaxation filters with double and multiple inversion,
  Magn.Reson.Imaging 25, 529-532 (2007). DOI: 10.1016/j.mri.2006.12.001. See also the Draft PDF.
• Bortolotti V., Fantazzini P., Gombia M., Greco D., Rinaldin G., Sykora S.,
  PERFIDI filters to suppress and/or quantify relaxation time components in multi-component systems: An example for fat-water systems; J.Magn.Reson., 206, 219-226 (2010). DOI: 10.1016/j.jmr.2010.07.010. See also the Draft PDF.
• More about PERFIDI on this site.
• DOI of this document: 10.3247/SL3Nmr10.004.