head and neck imaging Colin S. Poon, MD, PhD, FRCPC Adjunct Assistant Professor of Radiology
Case examples
Yale University School of Medicine U.S.A.
Head and Neck Imaging Skull base Extracranial neck Temporal bones Temporomandibular joints Pharynx / Larynx Cranial nerves Brachial plexus Vessels
Head and Neck Imaging CT MR PET
CT
CT Neck
Preferred technique for extracranial
structures in most institutions
Accessible Less prone to motion Patient motion (sick patient) Physiological motion (swallowing,
breathing) High spatial resolution Good contrast resolution
Importance of Multiplanar Reformats
Head and neck anatomy are complex
and pathology may be subtle Lesions may not be equally apparent
on all imaging planes. Multiplanar reformats are important for
head and neck imaging. Most useful planes are axial and coronal.
Puff Cheek Technique
45 year old male, adenocarcinoma of the hard palate
The lesion is easy to miss on axial images, but ismore obvious on the coronal reformatted images.
CT images with standard technique shows no obvious abnormality
CT images with “puffed cheek” technique, shows a buccal adenoid cystic carcinoma
Dual Energy CT
Dual Energy CT
Dual energy CT correctly disproves false positive laryngeal cartilage invasion.
Conventional CT
Dual Energy CT Kuno H, et al. Radiology 2012; 265: 488-496
55 year old Chinese man with conductive hearing loss
Dual energy CT correctly diagnose true positive laryngeal cartilage invasion.
Conventional CT
Dual Energy CT Kuno H, et al. Radiology 2012; 265: 488-496
Nasopharyngeal Carcinoma with Intracranial Extension
MRI for Head and Neck Superior soft tissue contrast Helps to detect subtle mucosal
abnormality Important for evaluation of skull base and cranial nerve lesion
MRI Best for imaging soft tissues, nerves Imaging protocol Thin sections (≤ 3 mm) Axial and coronal planes T1 and T2 weighting Includes post-gadolinium imaging To fat-suppress or NOT to fat-suppress, this is the controversy Air – bone – soft tissue interface results in severe
magnetic susceptibility that can lead to failure in fat suppression and image artifacts
T2W fat-sat
Contrast enhancement is not necessary for diagnosis if reliable fat suppression is available.
Glomus vagale - with extension into jugular foramen and cerebellopontine angle
Post-Gadolinium
Brachial plexus
Reliable Fat Suppresion is Important for Head and Neck Imaging
Applications of FatSuppression Head and neck imaging
The Basic MR Physics
Brachial plexus
Protons are like small magnets precessing along axis of main magnetic field
Fat and water have different precessing frequencies
Static magnetic field = B0 Leads to off resonance effect Radiofrequency magnetic field = B1 Leads to non-uniform RF power
SPAIR
SPAIR
Spectrally Adiabatic Inversion Recovery Spectrally = select fat only Adiabatic = special RF pulse insensitive to B1
inhomogeneity Inversion Recovery = null fat signal
Limitation: Does not solve the problem of B0 inhomogeneity
T2W
T2W SPAIR
Targeted Shimming Targeted Shimming No targeted shimming
Targeted shimming Targeted Shimming Volume
SPAIR After Targeted Shimming
Radiofrequency Energy Tips Magnetization Away From Longitudinal Axis of Main Magnetic Field Non-uniform RF power will lead to variation in fat saturation
Effect of B0 and B1 Inhomogeneity on Chemical Selective Fat Saturation Pulses A. Normal fat saturation
B. B0 inhomogeneity results in
incomplete fat suppression
C. B0 inhomgejeneity results in
inadvertent water suppression
B. B1 inhomogeneity results in
incomplete fat suppression
Fat Suppression
Inversion Recovery
STIR
Chemically Selective Presaturation
STIR
STIR Pros Robust to B0 and B1
inhomogeneities Reliable fat suppression
Cons Mixed constrast Inherent T1 weighting
Only works with proton
density and T2W Low SNR efficiency Suppresses short T1 species
and enhancing tissue after contrast Inversion of Magnetization
Wait until fat signal goes through null point
Chemically Selective Fat-Sat vs. STIR
Comparison of short inversion time inversion recovery (STIR) and fat-saturated (chemsat) techniques for background fat intensity suppression in cervical and thoracic MR imaging Nakatsu M. , et al. J. Magn. Reson. Imaging
2000; 11:56-60
Five-point rank score analyses were utilized by
Chem Fat-Sat
three experienced radiologists. The mean scores of STIR and fat-saturated FSE techniques for uniformity of fat suppression were 4.3 and 2.3, respectively (P < 0.0001). The mean scores of STIR and fat-saturated FSE techniques for lesion conspicuity were 4.2 and 3.5, respectively (P < 0.0001).
STIR
STIR vs. Chemical Saturation J. Magn. Reson. Imaging
STIR
Uniformity of Fat Suppression
2000; 11:56-60
Chem Sat FSE T2
FSE T2
Score
STIR FSE
Fat-saturated FSE
5
18
4
16
3
5
2
1
1
0
11
Total
40
40
Mean
4.3
2.3
0 7
7 15
Lesion Conspicuity Score
STIR FSE
5
17
3
4
10
20
3
7
10
2
1
1
1
0
1
35
35
Total Mean
Fat-saturated FSE
4.2
Chemical Shift Based WaterFat Separation Dixon Technique and Variations
3.5
Original Dixon Technique
Original Two Point Dixon Technique Sensitive to B0 inhomogeneity
Solution: In-phase = Water + Fat Out-of-phase = Water – Fat
Map the magnetic field inhomogeneity by acquiring additional points and applying correction in reconstruction
In-phase + Out-of-phase = Water only In-phase – Out-of-phase = Fat only
=> Three-point or four-point Dixon
T1 and T2weighted fast spinecho imaging of the brachial plexus and cervical spine with IDEAL water–fat separation
IDEAL Iterative Decomposition of Water and Fat with Echo Asymmetry and Least Squares Estimation
Recombined
Water
Fat
CHESS-FSE
IDEAL
Journal of Magnetic Resonance Imaging Volume 24, Issue 4, pages 825-832, 12 SEP 2006 DOI: 10.1002/jmri.20721 http://onlinelibrary.wiley.com/doi/10.1002/jmri.20721/full#fig1
Fat Suppression Techniques
Chemical Shift Water-Fat Separation Pros Robust to B0 and B1
inhomogeneities
Reliable fat suppression
Provides fat and water
Cons Long scan times More complex
reconstruction
Chem Fat-Sat
Method
Pros
Cons
Applications
Chemically selective fat suppression
• Relatively fast • Applicable to most pulse sequence
• Sensitive to B0 and B1 inhomogeneities
Poor performance for head and neck, brachial plexus
Spatial-spectral pulses
• Insensitive to B1 inhomogeneities
• Sensitive to B0 inhomogeneities
STIR
• Insensitive to B0 and B1 inhomogeneities • Reliable