Premium
Multilevel wavelet‐transform encoding in MRI
Author(s) -
Peters Robert D.,
Wood Michael L.
Publication year - 1996
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.1880060317
Subject(s) - wavelet , wavelet transform , discrete wavelet transform , second generation wavelet transform , wavelet packet decomposition , stationary wavelet transform , artificial intelligence , computer science , harmonic wavelet transform , pattern recognition (psychology) , fast wavelet transform , computer vision , fourier transform , encoding (memory) , mathematics , mathematical analysis
This paper explores an alternative to standard Fourier MRI called wavelet‐transform encoding. Spatially selective radiofrequency (RF) pulses are used to excite slice profiles in the shape of the wavelet‐basis functions of a discrete wavelet transform. Our implementation on a standard commercial whole body MRI system resolves one spatial dimension through wavelet encoding and orthogonal directions through conventional frequency encoding and rectangular‐slice selection. The wavelet transform is described as a method for multiresolution analysis, and we show how a high resolution MR image can be constructed from lower resolution images, representing the approximation and detail structures of the object. A multiple level architecture of the wavelet reconstruction is described, thereby allowing wavelet‐encoded images to be reconstructed through several modes. Multilevel wavelet‐encoded images of a gel phantom are presented that show comparable image quality to Fourier‐encoded images of similar signal‐to‐noise ratio. However, a chemical‐shift artifact is identified with this particular implementation.