Image quality and artifacts
Learning objectives
After reading this chapter, you should be able to:
- Present the different parameters used to judge the quality of an image
- Describe the factors influencing the signal-to-noise ratio and their interdependence
- List the different MRI artifacts, their origin, effects on the image and ways of reducing them:
- Movements, phantom images, flow
- Magnetic susceptibility and metal artifacts
- Truncation / Gibb’s
- Folding / aliasing
- Chemical shift
- Cross-excitation
- Magic angle
- Present the basic criteria in MRI quality control
Key points
Signal to noise ratio
Principles
Compromise between:
• Spatial resolution: limited by the voxel size which is determined by the matrix size, the field of view and slice thickness
• Signal to noise ratio: depending on the voxel size, the number of averagings and the receiver bandwidth
• Total scan time.
Which also modify the available sequence parameters (TE) and the artifacts.
Remedies (and penalties)
The signal to noise ratio goes up | Penalties |
With the voxel size | Limits the spatial resolution |
When a local or surface coil is used instead of a large or body coil | Reduced field of exploration |
With the number of averagings (square root factor) | Longer scan time |
When the receiver bandwidth decreases (square root factor) | Higher chemical shift artifacts |
Motion and ghosting
Principles
Motion is responsible for a corruption in spatial localization in the phase-encoding direction, resulting in a blur and ghost images propagated in the phase-encode direction.
Remedies (and penalties)
Remedies | Penalties |
Reduce body movements (physical restraint, sedation) | Patient's comfort |
Breath-hold sequences | Requires apnea and fast sequences |
Gating and movement compensation | Increase the scan time, restrict the available TRs, a blur persists with some methods |
Signal suppression of moving tissues (fat of abdominal wall...) | Increased TR due to magnetization preparation or saturation bands |
Swapping phase-encode and frequency-encode directions | Only shifts the artifacts, consequences on sequence parameters and other artifacts |
Magnetic susceptibility
Principles
At the interface between two tissues with different magnetic susceptibilities, there are local distortions in the magnetic field responsible for a signal loss (and sometimes an image distortion). These artifacts are much stronger in presence of metal.
Remedies (and penalties)
Remedies | Penalties |
SE rather than GE sequences | |
Short TE | Restricts available contrasts |
Increased receiver bandwidth to reduce the minimal TE | Decreased SNR |
Usage
• Detection of hematomas (blood breakdown products)
• Quantification of liver iron content
• Detection of liver metastases
• Perfusion imaging
Truncation
Principles
The image is reconstructed from k-space by a 2D inverse Fourier transform. As k-space data are finite and defined by the matrix size, the reconstruction of high-contrast interfaces is imperfect and causes visible artifacts. These artifacts appear as multiple parallel lines adjacent to the interface or as false widening of the edges at this interface. Truncation artifacts can occur in both the frequency and phase-encode directions.
Remedies (and penalties)
Remedies | Penalties |
Incread matrix size | Increased scan time Decreased SNR |
Aliasing
Principles
Aliasing or wrap-around results from a spatial mismapping caused by an undersampling in the phase-encode direction. Consequently, objects outside of the FOV overlap on the opposite side of the image.
Remedies (and penalties)
Remedies | Penalties |
Swapping the frequency-encode and phase-encode directions | |
Increasing the FOV | Decreased spatial resolution |
Phase oversampling | Increased scan time |
No-phase wrap or Anti-aliasing | Decreased SNR |
Chemical shift
Principles
The chemical environment of protons can cause a shift in their precessional frequency. The shift in resonance frequency between protons in fat and water is roughly equal to 3.5 ppm, corresponding to a difference of about 225 Hz at 1.5 T.
This frequency shift is responsible for a spatial misregistration in the frequency-encode direction, resulting in a contour artifact with dark and white bands at the interfaces between fat and tissues in the frequency-encode direction : the chemical shift artifact of the first kind.
As the resonance frequency shift causes a phase shift which is not canceled in gradient echo sequences, there is a black line at all fat/tissue borders when fat and water are out of phase (TE = 2.2 ms at 1.5 T) : the chemical shift artifact of the second kind. This artifact occurs in both the frequency-encode and phase-encode directions.
Remedies (and penalties)
Remedies | Penalties |
Fat suppression | Increased scan time |
Swapping of the frequency-encode and the phase-encode directions | Rotates the artifact without eliminating it |
Increased receiver bandwidth | Decreased SNR |
Cross-talk
Principles
• Imperfect slice profile or intersecting slice stacks
• Contrast modification and/or signal loss
• Mainly with multislice spin echo technique, fast spin echo and inversion-recovery sequences
Remedies (and penalties)
• Gap between slices
• Interlacing
Magic-angle
Principles
• Dipolar interactions in fibrillar structures, varying according to the angle of the fibers in relation to the axis of field B0.
• Minimum at 55°: T2 relaxation time lengthening and increased T2-weighted signal intensity.
Remedies
• T1-weighted sequences
• Long TE
• Modify angle between fibrillar structure and B0 axis
Usage
Tendons and ligaments imaging (T1-weighted, +/- gado / MTC)
Quality control
Principles
• Signal: SNR, uniformity
• Geometry: linearity and distortion, spatial resolution and encoding
• NMR: T1 and T2 contrast, CNR, accuracy, precision
• Artifacts
• Spectroscopy: B0 homogeneity, SNR of main peaks
References
- Elster. Questions and answers in magnetic resonance imaging. 1994:ix, 278 p.
- McRobbie. MRI from picture to proton. 2003:xi, 359 p.
- NessAiver. All you really need to know about MRI physics. 1997.
- Kastler. Comprendre l'IRM. 2006.
- Korin, Felmlee. Adaptive technique for three-dimensional MR imaging of moving structures. Radiology. 1990 Oct;177(1):217-21.
- Harris and White. Metal artifact reduction in musculoskeletal magnetic resonance imaging. The Orthopedic clinics of North America. 2006 Jul;37(3):349-59, vi.
- Hood, Ho. Chemical shift: the artifact and clinical tool revisited. Radiographics. 1999 Mar-Apr;19(2):357-71.
- Bydder, Rahal. The magic angle effect: a source of artifact, determinant of image contrast, and technique for imaging. J Magn Reson Imaging. 2007 Feb;25(2):290-300.
- de Certaines and Cathelineau. Safety aspects and quality assessment in MRI and MRS: a challenge for health care systems in Europe. J Magn Reson Imaging. 2001 Apr;13(4):632-8.
- Ihalainen, Sipila. MRI quality control: six imagers studied using eleven unified image quality parameters. European radiology. 2004 Oct;14(10):1859-65.
- Zhuo and Gullapalli. AAPM/RSNA physics tutorial for residents: MR artifacts, safety, and quality control. Radiographics. 2006 Jan-Feb;26(1):275-97.