Noura Azzabou1,2, Paulo Loureiro De Sousa3,4,
Pierre G. Carlier4,5
1NMR Laboratory, ,
Institute of Myology, Paris, France; 2NMR Laboratory, CEA, I2BM,
MIRCen, IdM,, Paris, France; 3NMR Laboratory, Institute of
Myology, Paris, France; 4NMR Laboratory, CEA, I2BM, MIRCen, IdM,
Paris, France; 5NMR Laboratory, Institute of Myology, Paris,
France
We focused here on B1 receive non-uniformity artifacts correction. We presented the techniques that can be used in the context of quantitative musculoskeletal NMR imaging. To estimate the inhomogeneity function, these approaches rely on the assumption of the signal uniformity inside the subcutaneous fat region. They include normalized convolution based techniques and those that approximate functions through cosine basis or Legendre polynomial sum. The comparison between them was achieved on data acquired using 3pt Dixon sequence and different types of coils. Experimental results showed that techniques based on Legendre polynomial or cosine basis provided better correction quality.