Chemical shift artifacts
Two types of chemical shift artifacts exist :
- Type 1 is seen in the frequency-encoding direction and only concerns field strengths higher than 1 T
- Type 2 can be found at any field strength but requires GE sequences with particular TEs.
The chemical environment of protons can cause a shift in their precessional frequency due to magnetic shielding provided by the electron shell.
This shift in resonance frequency exists between protons in fat and water. It is roughly equal to 3.5 ppm, corresponding to a difference of about 225 Hz at 1.5 T.
The chemical shift artifact of the first kind results from a shift in the spatial location of fat voxels. The spatial location in the frequency-encode direction is assumed to be the consequence of the frequency-encoding gradient. Due to the chemical shift, voxels containing fat will not have the expected resonance frequency and will be spatially misregistrered, causing a shift in the spatial location in the frequency-encode direction. This artifact appears as a dark (or white) band at the interfaces between water and fat.
For a given field strength, the resonance frequency shift between protons of fat and water is constant (225 Hz at 1.5 T). The resulting chemical shift artifact is affected by the receiver bandwidth.
The narrower is the receiver bandwidth, the narrower is the bandwidth per pixel is and the wider the chemical shift in number of pixels is.
The chemical shift artifact of the second kind only occurs with gradient echo sequences. With spin echo sequences, the 180° pulse refocuses spins to create the echo. The absence of a 180° RF pulse in gradient echo sequences causes a phase shift between protons of fat and water when the (gradient) echo is formed. This phase shift depends on their resonance frequency shift due to the chemical shift. With a 1.5 T field strength, the frequency shift is 225 Hz, corresponding to a period of 4.4 ms. Therefore, at 1.5 T, protons of fat and water will be in phase every 4.4 ms : their signals are additive. For TEs between this interval of 4.4 ms, their phases are shifted and for a TE at the middle of this interval (2.2 ms), they are out of phase.
The signal intensity of a voxel containing fat and water oscillates with an increasing echo time, with a minimum when fat and water are out of phase, and a maximum when they are in phase.
for TE corresponding to fat and water out of phase (2.2 msec, 6.6 msec etc...), the signal of voxels containing the same proportion of fat and water is canceled, producing a black line at all fat/tissue borders. This contour artifact is known as the chemical shift artifact of the second kind.
This artifact occurs in both the frequency-encode and phase-encode directions as it is independent of spatial encoding.
It is never seen with spin echo sequences as the phase shifts due to chemical shift are canceled by the 180° refocusing pulse
The chemical shift artifacts are reduced by fat suppression techniques (saturation, inversion-recovery). The reduced signal from fat thereby minimizes the chemical shift artifact.
By swapping the frequency-encode and phase-encode directions, the chemical shift artifact is rotated in the other direction, without eliminating it.
This swapping also modifies the other artifacts (wrap-around, motion artifacts,...) and sequence parameters.
Another method is to use a wider receiver bandwidth: the wider the receiver bandwidth is, the wider the bandwidth per pixel is, and the less visible the chemical shift artifact will be. The penalty is a decreased signal to noise ratio.
One usage of the chemical shift artifact of the second kind is tissue-characterization. The comparison of tissue signals between gradient echo sequences in phase and out of phase allows for the identification of tissues containing a mix of fat and water, which is useful for the diagnosis of focal fatty liver or adrenal adenomas.
Increasing interventional radiology capacity while reducing patient radiation
Ensuring performance of x-Ray equipment: a holistic approach
A healthy dose of radiation monitoring
The first fully digital C-arm
Patient-Specific Radiation Dose Estimation in Breast Cancer Screening