These notes deal with technical issues governing sequence choice in upper abdominal imaging, the appearance of key pathologies, suggested protocols and associated references.
 

Technical Challenges	Generic Protocols
Current Solutions	Liver   Pancreas   MRCP
Breath Hold Sequence Consideration  T1 Weighted Techniques   T2 Weighted Techniques   Breath Hold Scans - Getting the Best Results	Imaging Characteristics of Liver Mass Lesions
Sequence Options	Stages of Enhancement
Respiratory Artefact Reduction   Flow Artefact Reduction   Gut Motion Artefact Reduction   Fat Suppression	Readings & References

• Artefacts from :-
• Respiratory motion ghosts
• Gut motion
• Flow related enhancement and pulsatile ghosts
• Chemical shift artefacts
• Achieving acceptable SNR at reasonable resolution
• Lesion characterization & sensitivity requires multiple contrast weightings, and multiple post contrast acquisition
• Achieving a practical examination time

Breath Hold Sequences Considerations

• Use breath holding sequences wherever possible. This approach gives faster examination times, sharper lesion borders.
• The stage of respiration must be consistent throughout the examination to avoid positioning errors and lost tissue in interleaved sequence pairs.
• Breath hold sequences need a phased array body coil
• For a sequence time <20 seconds then,  <20

• (ETL = EPI factor + Turbo factor)
• T1 imaging 128 phase steps & NEX = 1 TR < 156 msec
• For T2 imaging TR > 2500 ETL /NEX > 8

(as TR increases required ETL increases)

• Sequence intended to identify focal lesions and for use post contrast to display rapid changes in enhancement.
• Short TR spoiled gradient echo (SGE) sequences FLASH, SPGR give best results.
• Combination of short TR (<160) and large number of slices (15 - 20) can restrict choice of sequence options that consume sequence time (fat suppression, spatial suppression, GMR)
• Most scanners should be able to provide a workable sequence. If slice numbers are restricted perform two interleaved sequences.
• Current & development sequences employ interleaved excitations, restricted application of fat sat routines or water excitation, no GMR, and restricted application of spatial saturation routines.

• Sequence used to identify regions of fatty infiltration and help identify adenomas.
• aka chemical shift imaging
• Relies of differences in the precessional frequency of Hydrogen in fat and water molecules.
• Gradient echo image with TE set to a time where fat and water spins are 180 degrees out of phase causing fat and water signal components within a voxel to destructively interfere and reduce the brightness of the pixel.
• Degree of signal change depends on the ration of fat/water. All fat or all water, no signal drop (compared to in-phase image) 50/50 causes total signal loss
• OOP images identified by dark bands at fat water interfaces
• Select the shortest OOP TE to get best T1 contrast and largest number of slices
• Exact TE depends on filed strength

T1 In Phase S-GRE (FLASH)  Te = 4 msec	T1 Out of phase S-GRE (FLASH)  Te = 2

T1 Fat Suppressed Intended to distinguish bright lesions on T1 images (blood from fat) and to improve display of the pancreas (especially when atrophic) Standard spectral saturation routines are too time consuming (10 msec/slice) Water excitation (non spatial, spectral selective) works well
	T1Breath hold FLASH with interleaved Fatsat

T2 Weighted Techniques

Breath Hold T2 TSE (RARE, FSE) Higher ETL allows use of long TR and long Te for very strong T2 weighting. Primary aim is discrimination between fluid filled lesions and solid tumours. The MTC effect of multiple refocusing pulses decreases contrast between liver and solid tumours. GRASE or multi-shot EPI sequences should minimize this effect T2 images should have fat suppression, especially for the pancreas.

T2 breathhold Fatsat TSE

T2 non-breath hold TSE Allows very high spatial resolution with phased array coils. Fat suppression and spatial saturation largely eliminate artefacts Can use lower ETL to reduce MTC effects and better discriminate solid lesions. (But they won't classify the lesion so this is a minor point)
	T2 TSE non breathhold

Long TE STIR Fat suppression is extremely uniform and robust Appearances virtually identical to a fat suppressed T2 TSE Double echo sequences offer a reverse T1 fat suppressed image as well  Usually a non-breath hold sequence
	Non-breathhold TSE STIR   Te 85 msec TI 160 msec

• Coach the patient with the breath hold technique before they go into the magnet. Maintain the same voice and instructions for all breath holds.
• Give the patient the headphone earmuffs
• Turn off the radio during the BH procedures and set the intercom to listen to the patient.
• Scan on suspended expiration. Ask the patient to take a comfortable breath in and a relaxed breath out then stop.:-
• breath in, breath out, and stop breathing
• Tell the patient they can breath again when the scan noises stop.
• Get the sequence set, then click LOAD.
• When the message LOADED appears in the measurement dialogue area give the breathing instructions while watching for abdominal motion on the monitor or directly.
• Start the sequence with the START button on the console keyboard and listen to the scan.
• Listen to the scan running

Respiration Artefact Reduction

 ROPE

• Respiratory ordered phase encoding collects centre of K-space when respiratory motion is minimal.
• Moderate scan time penalty (25%) & not entirely effective.
• Can be useful

Respiratory Gating

• Data collection restricted to periods of minimal respiratory motion (excitation continues).
• Moderately effective.
• High scan time penalty (50%-100%) & not very effective.
• Increases SAR substantially.
• Essentially useless

Respiratory Triggering

• Machine based triggering or user selected TR to match breathing period.
• Usable for long TR sequences only
• Quite effective but not robust

Multiple Acquisitions

Average motion artefacts over multiple acquisitions

Typically only works with NEX >4

[repeat phase step n times then increment, average over TR x NEX]

Better with "SMART" averaging (Philips)

(collect all phase steps then repeat acquisition, average over TR x N.phase x NEX)

Only moderately effective

Increased scan time and opportunity for other patient motion

Breath Hold Sequences

• Restrict sequence period to <25 seconds
• Extremely effective (no respiratory artefact or repeat the scan)
• Presents difficulties for sequence design and parameter choices.
• Practical application depends on scanner performance specifications.
• Gives sharper images by avoiding motion induced partial volume averaging effects

Ultrafast Acquisition

• Sequential acquisitions of single slice sequences running <1 second
• Obviates need for breath hold (better with it)
• Overcomes some limitations of breath hold sequences
• Practical with high performance equipment

Standard Liver
To detect and characterize focal liver lesions

• Scout (multiple planes)
• T2 weighted Axial with fat suppression
• T1 out of phase Spoiled Gradient Echo (SGE) Axial
• If the lesion is bright on T1 do a fat saturated T1 Axial
• T1 in phase SGE Axial

• T1 In Phase FLASH Axial at :-
• End of saline flush,
• 45 seconds post contrast
• 3 minutes post contrast
• 10 minutes post contrast (optional)
• MRCP axial, RAO for Hepatic ducts, LAO for pancreatic duct

• The pre-contrast images locate lesions and provide T1 and T2 weighted information.
• Out of Phase (OOP) FLASH images display fatty infiltration and adrenal adenomas.
• The post contrast images help classify the nature of the lesion.
• The immediate post contrast image is very time critical, and is the most valuable.
• The rapid injection requires good venous access. Have the 20 G Jelco in situ before the patient is placed in the bore ( before they enter the room if possible).

To detect and characterize focal pancreatic lesions

• Scout (multiple planes)
• T2 weighted Axial with fat suppression
• T1 out of phase Spoiled Gradient Echo (SGE) Axial
• T1 in phase SGE Axial with fat suppression

• T1 In Phase FLASH Axial with fat suppression at :-
• End of saline flush
• 45 seconds post contrast
• 1.5 minutes post contrast
• MRCP Thick slice axial, RAO for Hepatic ducts, LAO for pancreatic duct

• An imaging technique that aims to replicate the information provided by endoscopic cholangiopancreatography (ERCP) and other cholangiographic techniques by developing very high contrast between fluid (in the ductal anatomy) and background tissues.
• Best results with extremely T2 weighted long ETL TSE sequences with fat suppression, acquired in a single breath hold.
• Slices < 4 mm can be processed in a MIP programme to allow some variation of projection
• Thick slices can provide very fast sequences with a single view and high in-plane resolution

Thick Slice Localiser or simple display of ductal anatomy:  Fat saturated single shot HASTE Axial and oblique coronal planes  TR 2800 TE 1100 ETL 128 Slice thickness 70 mm FOV 300 mm Matrix 256 x 240 Resolution 1.27 x 1.2 mm Acquisition time 7 seconds ( uses 2 TR to establish steady state)
Thin Slice - For detailed depiction of gallstones or ductal obstruction  Fat suppressed single shot HASTE Coronal  TR 11.9 (infinite) Te 95 ETL 128 13 slices 4 mm thick FOV 270 mm (7/8) matrix 256 x 240 resolution 1.13 x 1.05 mm Acquisition time 20 seconds

• Capillary Phase a.k.a. Early enhancement or Non Selective Hepatic Arterial.
• Characterized by contrast in the hepatic arteries but not in the portal or hepatic veins
• Early hepatic venous Phase. a.k.a. early non equilibrium
• Contrast in all hepatic vessels
• Occurs less than 1 minute post contrast
• Equilibrium Phase
• 1 to 5 minutes post contrast
• Washout Phase
• 10 minutes, or more, post contrast

 

Early enhancement & Washout  Hepatocellular Carcinoma Hepatic Adenoma Cholangiocarcinoma Hypervascular metastases (unusual)	Bright on T2 Haemangioma Hepatocellular Carcinoma Cholangiocarcinoma Metastases Simple Cysts (very bright esp on Te > 120)
Contains Fat or blood  Hepato-cellular Carcinoma Adenoma (most drop signal on OOP image)	Pseudo-capsule (compressed liver) Hepato-cellular Carcinoma Adenoma
Decreased Signal on OOP image Adenoma Fatty infiltration (cirrhosis, post chemotherapy)	Uniform enhancement Fibro-nodular Hyperplasia Adenoma
Persistent rim enhancement metastases Late enhancement & delayed washout Haemangioma very rarely hypervascular metastases	Rim enhancement and peripheral washout hypervascular metastases

• Clinical MRI 2nd Edition Edelman Hesselink & Zlatkin Volume 2 Page 1467
• Fast MR Imaging Techniques : Impact in the Abdomen. D.G. Mitchell JMRI 1996 6:812-821
• Recent Technical Advances in MR Imaging of the Abdomen SJ Riederer JMRI 1996 6;822-832
• MR of Focal Liver Lesions: Comparison of Breath Hold and Non-Breath Hold Hybrid RARE and Conventional Spin-Echo T2 Weighted Pulse Sequences K.D. Carpenter et al JMRI 1996 6:596-603
• HASTE MR Imaging: Description of Technique and Preliminary Results in the Abdomen R.C. Semelka N.L. Kelekis, D. Thomasson, M. Brown, G.Laub. JMRI 1996 6:698-699
• Technical factors in MRI Body Imaging. Richard Semelka Book of Abstracts The Inaugural Gold coast MRI & CT Conference 1996
• Liver. Richard Semelka Book of Abstracts The Inaugural Gold coast MRI & CT Conference 1996
• Pancreas. Richard Semelka Book of Abstracts The Inaugural Gold coast MRI & CT Conference 1996
• Body MRI. Richard Semelka Book of Abstracts The Inaugural Gold coast MRI & CT Conference 1996

Click Me to Mail to Me If you would like this paper presented at your MRI meeting or you have comments or suggestions, please contact the author by e-mail by clicking on the picture.