2025 SPR实践指南：儿童肝脏MRI检查

Abstract

Magnetic resonance imaging (MRI) has become an essential tool in the evaluation of pediatric liver disease. However,the unique physiological, anatomical, and behavioral characteristics of pediatric patients present distinct challenges that necessitate tailored imaging strategies. These guidelines, developed by members of the Society for Pediatric Radiology(SPR) Magnetic Resonance and Abdominal Imaging Committees, provide comprehensive recommendations for perform-ing high-quality liver MRI in children. Drawing on multidisciplinary expertise from pediatric radiologists, MRI physicists,and technologists across diverse institutions, the document addresses key technical considerations, including field strength selection, motion mitigation, contrast agent use, and age-specific protocol optimization.These guidelines aim to standard-ize and elevate the quality of pediatric liver MRI, offering practical, evidence-informed recommendations to support safe,efficient, and diagnostically robust imaging across a wide range of clinical scenarios.

Keywords Liver·Pediatric·SPR

Introduction

The use of MRI to evaluate congenital and acquired liver diseases in children is the standard of care for many clinical indications, including the evaluation of biliary abnormalities[1], vascular anomalies[2], liver tumors[3], and quantitative liver imaging[4]. These diseases occur from the neonatal period to the teenage years, encompassing children of a wide spectrum of sizes and varying ability to follow commands.Many routine abdominal MR imaging sequences require breath-holds, which generate challenges in both awake and sedated/anesthetized pediatric patients. These challenges require either adapting existing sequences for use during free-breathing or utilizing sequences specifically designed to compensate for respiratory motion. Additionally, simple changes in parameters(e.g., decreasing the field of view for infants) modify many other parameters as a cascade in the background, leading to struggles with real-time image quality optimization in children. It is helpful to provide tech-nologists with pre-optimized protocols tailored to children of different sizes and breath-holding abilities. Given the unique technical and diagnostic challenges of MR evaluation of infants and children with abdominal diseases, there has

Preamble

The Society for Pediatric Radiology(SPR) Magnetic Resonance Imaging(MRI) Committee was established to advance the quality, safety, and availability of MRI for children. The SPR Abdominal Imaging Committee was established to advance the application of abdominal imaging techniques, including MRI, in children. Through ongoing research and education initiatives, these committees diligently uphold the highest standards of excellence in pediatric MRI, ensuring optimal outcomes for young patients. Leveraging its collective expertise,these committees promote knowledge sharing, facilitate interdisciplinary dialogue, and advocate for the integration of cutting-edge technologies in pediatric MRI.These guidelines, endorsed by the SPR MRI and Abdominal Committees, serve as educational resources to assist imagers in delivering optimal care to pediatric patients. While they provide valuable insights, they are not inflexible rules or legal standards.The SPR MRI and Abdominal Imaging Committees underscore the need for individualized patient care decisions based on current knowledge, resources, and patient needs, aiming to ensure effective and safe medical practices in pediatric MRI.

Extended author information available on the last page of the article

Pediatric Radiology

been interest in creating guideline documents for imaging pediatric patients. Herein, we provide guidelines and imag-ing suggestions to address common imaging challenges in children, including the standard imaging sequences obtained for the purpose of liver evaluation.

These guidelines and imaging suggestions reflect a con-certed effort of members of the SPR MRI and Abdominal Imaging Committees who have specific expertise in pediat-ric liver imaging, clinical MRI physicists, and pediatric MRI technologists, all from children's imaging programs of vari-ous sizes with a variety of common MRI vendor platforms.The aim of these guidelines is to provide practical guid-ance for the implementation and application of up-to-date MR imaging techniques in all pediatric patients. The final document was reviewed and approved by the full MRI and Abdominal Imaging Committees and ultimately approved by the SPR Board.

# Field strength considerations

Pediatric liver MRI is typically performed at 1.5 T or greater.The advantages and limitations of different field strengths are noted below.

1. Higher field strengths(e.g., 3 T) offer improved signal-to-noise ratio(SNR), enabling higher resolution and/or faster imaging.

2.Higher field strengths tend to have poorer magnetic field homogeneity, which can make chemical faturation difficult to achieve consistently.

(a) Alternative fat suppression methods, such as STIR or Dixon techniques, are less sensitive to inhomo-geneities.

3. Reduced magnetic field homogeneity also makes geo-metric distortions more severe, which is particularly problematic in diffusion-weighted imaging(DWI).

(a) Decreasing echo spacing by increasing receiver bandwidth, using parallel imaging acceleration,or performing multi-shot diffusion imaging can reduce distortion.

4.Susceptibility artifacts are more prevalent at higher field strengths. Using shorter echo times or a higher receiver bandwidth is an effective strategy to manage these arti-facts [5].

5.MRI at 1.5 T may be preferred for liver iron quantifica-tion since T2* decay is slower at 1.5 T, enabling a larger range of echo times to be employed with sufficient SNR for iron quantification. Additionally, more calibration data is available at 1.5 T than 3 T[5].

<BeginOfPara>6. MRI safety concerns are typically more restrictive at higher field strengths because of the increased radiof-requency(RF) energy deposition and increased specific absorption rate(SAR). In addition, devices and implants

that are MR conditional at one field strength may be unsafe at a different field strength. MRI conditional-ity statements should always be reviewed for each field strength used for imaging.

# Specialized imaging equipment

1. RF coils should be selected based on patient size and they should be oriented to maximize the benefits of acceleration.

(a) Coil arrays with sizes closely matching the desired field of view of view FOV) will offer improved image SNR.

(b) Dedicated pediatric coils are offered by multiple MRI vendors and third-party vendors[6].

(c) Adult coil arrays with high element densities, such as cardiac arrays, head/neck vascular coils, and flexible extremity coils, are good alternatives to dedicated pediatric coils depending on patient size[7].

(d) Multichannel arrays enable parallel imaging for accelerated acquisitions. Larger channel count coils offer greater potential for image accelera-tion. However, the advantages of greater imaging speed must be balanced with the inherent SNR loss from parallel imaging.

2.MR elastography exams require additional phase-sensi-tive pulse sequences, an elastography driver, and post-processing software.

3.Diffusion-weighted imaging benefits from high-perfor-mance gradients, though they are not a requirement.

4.Iron and fat quantification typically requires additional multi-echo pulse sequences and post-processing soft-ware.

Gadolinium-based contrast agents

Pediatric liver MRI routinely uses the paramagnetic prop-erties of gadolinium-based contrast agents(GBCAs) to enhance lesion detection and characterization. However,group I classified GBCAs add a potential risk of nephro-genic systemic fibrosis(NSF) in patients with stage 3 or worse renal failure[8]. There is also evidence of gadolinium accumulation in other tissues[9]. To reduce potential toxic-ity, gadolinium is chelated with other compounds, which are typically described by their relaxivity and are classified by the American College of Radiology(ACR) according to their stability.

The eight current FDA-approved GBCAs for pediatrics are listed in Table 1. GBCA safety and administration con-siderations are noted below.

Pediatric Radiology

Table 1 Current FDA-approved gadolinium-based contrast agents approved for pediatric use

|Contrast agent|Trade name|Type|T1 relaxivity in human plasma at 37°C $${M}^{-1}\,s^{-1}$$ (1.5 T/3) *|T2 relaxivity in human plasma at 37°C, LmM ${}^{-1}$ s ${}^{-1}$ (1.5 T/3) *|Clearance|ACR classifica-tion|FDA-approved ages| |---|---|---|---|---|---|---|---| |Gadodiamide|Omniscan|Linear|4.3/4.0|5.2/5.6|Renal|Group I|≥2 years| |Gadopentetate dimeglumine|Magnevist|Non-ionic Linear|4.1/3.7|4.6/5.2|Renal|Group I|≥2 years| |Gadobentate dimeglumine||Non-ionic Linear|6.3/5.5|8.7/11.0|Renal (95-96%) and hepa-tobility (4-5%)|Group II|All ages| |Gadobutrol|Gadavist/ Gadovist|Non-ionic Macrocyclic non-ionic|5.2/5.0|6.1/7.1|Renal|Group II|All ages| |Gadoteric acid|Dotarem|Macrocyclic ionic|3.6/3/5|4.3/4.9|Renal|Group II|All ages| |Gadoteridol|ProHance|Macrocyclic non-ionic|4.1/3.7|5.0/5.7|Renal|Group II|All ages| |Gadoxetate disodium|Eovist/|Linear|6.9/6.2|8.7/11.0|Hepatobiliary (50%) and renal (50%)|Group III|All ages| |Gadopicenol|Primovist Vueway/Elu-cirem|Ionic Macrocyclic non-ionic|19/9.9|34/60|Renal|Group II|≥2 years|

Information in this table is based on the following references:

Robic C, Port M, Rousseau X, Louguet S, Fretellier N, Catoen S, Factor C, Le Greneur S, Medina C, Bourrinet P, Raynal I, Idee JM, Corot C,Physicochemical and pharmacokinetic profiles of gadopilenol: a new macrocyclic gadolinium chelate with high T1 relaxivity. Invest Radiol.2019 Aug;54(8):475-484. PMCC:6661244

Shen Y, Goerner FL, Snyder C, Morelli JN, Hao D, Hu D, Li X, Runge VM. T1 relaxivities of gadolinium-based magnetic resonance contrast agents in human whole blood at 1.5, 3, and 7 T. Invest Radiol. Ovid Technologies(Wolters Kluwer Health); 2015 May;50(5):330-338. PMID:25,658,049

Rohrer M, Bauer H, Mintorovitch J, Requardt M, Weinmann HJ. Comparison of magnetic properties of MRI contrast media solutions at different magnetic field strengths. Invest Radiol. Ovid Technologies(Wolters Kluwer Health); 2005 Nov;40(11):715-724. PMID: 16,230,904

1.Non-contrast techniques should be considered alterna-tives for contrast-enhanced techniques whenever pos-sible.

2.Special consideration to minimize gadolinium dose should be based on the child's measured weight and renal function in addition to GBCA relaxivity and stability.

(a) Higher relaxivity GBCAs, such as gadopiclenol,offer the ability to use lower gadolinium dosages.

(b)Macrocyclic compounds are the most stable class of GBCAs and pose the lowest risk of NSF.

3.Patients who have elevated total bilirubin and/or evi-dence of liver failure may have poor excretion of gadox-etate into the biliary tree, rendering MRCP evaluation limited or non-diagnostic[10]. Assessment of these fac-tors prior to examination is recommended.

Motion management

Pediatric MRI has unique considerations that influence motion management strategies. Coaching patients to perform breath-holds and remain still is less successful in younger pediatric populations. Consequently, prioritizing patient comfort, employing effective motion mitigation strategies,utilizing rapid imaging techniques, and incorporating res-piratory gating become crucial elements in achieving diag-nostic image quality. Considerations for pediatric motion management are noted below.

Patient comfort and preparation

1. Patient cooperation with lying still and performing breath-holds can be improved by reducing patient anxi-ety [11].

Pediatric Radiology

(a) Child life specialists provide education, support,and comfort for children and their families before and during their imaging procedure.

(b) Practicing breath-holds before imaging may help patients get used to the required breath-hold durations and inform the technologist about their breath-holding capabilities.

(c)Decorating MRI rooms and equipment with child-friendly themes reduces anxiety.

(d) Dedicated realized or virtual mock MRI environments that simulate the equipment, sounds, and vibra-tions prepare children for what to expect during an MRI exam.

(e) Audiovisual equipment captures the child's atten-tion during the procedure, fostering compliance.

(f) Hearing protection, including ear plugs and head-phones, reduces the potential for permanent hear-ing loss and also prevents motion induced by star-tling MRI sounds. Many vendors now also offer quieter gradient modes, which can be helpful to reduce anxiety.

2. For infants up to 6 months of age, feeding and bundling are effective at reducing bulk patient motion, but do not improve respiratory motion artifacts[12].

3.For patients more than 6 years old, breath-hold imag-ing may be possible with appropriate coaching. Breath-holds are typically restricted to 16 s or less, which con-strains the achievable spatial resolution and coverage.

4.For young pediatric patients, between 6 months and 6 years old, sedation or general anesthesia may be needed to suppress patient motion to achieve diagnos-tically useful images. A risk-benefit analysis must be incorporated into the decision whether to use sedation or anesthesia or rely on other motion mitigation strategies.The depth and length of anesthesia should be consid-ered, as breath-holds may require deeper anesthesia[6]

Motion mitigation sequences and strategies

1.The effects of periodic translational motion can be reduced by frequently sampling the center of k-space.This can be accomplished with sampling strategies that use radial blades of parallel lines(e.g., BLADE-Sie-mens Healthineers, PROPELLER-GE HealthCare,MultiVane XD-Philips Healthcare) or radial spokes(GRASP, StarVIBE-Siemens; Disco Star, LAVA Star-GE Healthcare; 4D Free Breathing, 3D Vane XD-Philips). Oversampling the k-space center can increase scan duration, so most clinical implementations mitigate this time penalty via parallel imaging or compressed-sensing reconstruction.

(a) Through-plane motion is not improved by two-dimensional sequences that use conventional Car-tesian trajectories in the slice direction[13].

2. Signal averaging can be used to reduce the conspicuity of motion artifacts in MR images.

(a) Additional averages lengthen the total image acquisition time[13].

(b) Phase oversampling, rather than full signal aver-ages, can also be used to mitigate motion with less impact on imaging time[13].

3. Swapping the phase-encoding direction can confine the motion artifacts to regions that do not overlap the liver.Caution may be warranted as phase artifacts may result and mimic pathology.

4. Employing fat suppression or saturation bands on the subcutaneous fat in the anterior abdomen wall can help to reduce motion artifacts from the hyperintense periph-eral fat in the moving anterior abdominal wall.

Rapid imaging and acceleration sequences and strategies

1. Single-shot and multi-shot imaging acquire multiple lines of k-space data for each TR, accelerating image acquisition at the expense of the caveats noted below.

(a) Longer echOS trains reduce contrast resolution, mix contrast weighting, and can cause $T_{2}$ -blurring[13].

(b) Additional RF pulses in single- or multi-shot spin echo imaging increase SAR.

2.Parallel imaging uses multiple receive coil elements to acquire fewer phase encodes, accelerating imaging 2-fourfold depending on the FOV, receive coil design, and channel count[14].

(a) There are two strategies of parallel imaging: image-based(SENSE-Siemens,ASSET-GE Healthcare,SENSE-Philips) and k-space-based(GRAPPA-Siemens, ARC-GE Healthcare). K-space-based parallel imaging typically manifests fewer artifacts for abdominal imaging, but using either parallel imaging strategy is highly recommended to reduce scan time and breath-hold duration[6].

(b) CAIPIRINHA(Siemens), an extension to 3D GRAPPA, can offer additional acceleration in pediatric liver MRI[13]

3. Partial Fourier imaging(partial Fourier-Siemens,Fractional NEX-GE Healthcare, Half Scan-Philips),a technique that exploits symmetries in k-space, is useful to further reduce scan time and breath-hold duration.

(a) Partial Fourier imaging offers a maximum two-fold acceleration; although practically, image

Pediatric Radiology

acquisition is expedited by less than two-fold to preserve image quality[15].

(b)Greater time savings are achievable when partial Fourier is used in the phase-encoding direction rather than the readout direction [6].

(c) Further acceleration can be achieved by using par-tial Fourier in conjunction with parallel imaging.

4.Compressed sensing and deep learning can now be exploited on modern clinical MRI scanners to further reduce scan and breath-hold times, offering two- to five-fold acceleration[14, 16].

(a) Image reconstruction with these techniques is slower, due to its iterative nature, and inherently results in image filtering.

(b) Physician feedback is needed to balance accelera-tion and filtration.

(i) High acceleration rates can lead to overly smooth images[6].

(c)Compressed sensing and deep learning are typi-cally used with parallel imaging to further accel-erate image acquisition. A careful and informed approach is advised with the application of deep learning reconstructions in pediatric MR[17].

5.Simultaneous multi-slice(SMS) imaging expedites imaging by concurrently acquiring multiple slices(simultaneous multi-slice-Siemens, Hyperband-GE Healthcare, MultiBand SENSE-Philips)

(a) Acceleration factors on the order of two-fold are achievable with SMS[13].

(b)SMS is compatible with parallel imaging in-plane.

Respiratory synchronization

Respiratory motion can be monitored in real-time to restrict imaging to specific phases of respiration. Respiration can be monitored using external triggers(e.g., respiratory bellows)or internal triggers(e.g., navigator echoes.)

1. Navigator echoes more accurately track diaphragmatic motion, capturing shallow and irregular breathing more reliably than bellows [13].

2.Respiratory gating techniques can increase scan time by 2-threefold or more, depending on the breathing pattern[13].

3.Respiratory gating using adaptive intelligence is avail-able on some modern scanners. This allows for accurate triggering without the need for respiratory bellows or an additional navigator pulse (e.g., Philips VitalEye).

Age-specific protocol optimization

The large variability in pediatric patient sizes and their abil-ity to breath-hold complicates protocol optimization and makes the use of a single protocol for all patients inefficient.

1. Pediatric patients have smaller anatomical features than adults, requiring higher spatial resolution imaging that could result in unnecessarily long scan times for a one-size-fits-all FOV.

2. Pediatric patients of different ages also have differing abilities to perform a breath-hold and remain still during scanning, influencing the selection of motion mitigation strategies.

3.It is recommended to have at least two, and preferably three, pre-optimized protocols for children of different sizes and abilities to breath-hold on your scanner for technologists to choose from. Suggested protocol sets for different pediatric patient ages are listed in Table 2 along with typical FOVs.

General techniques

MRI of the liver is performed for a wide spectrum of focal and diffuse liver diseases in children. Typical MR protocols will include core sequences performed in most cases to evalu-ate liver anatomy and parenchyma, with the addition of spe-cific sequences tailored for the indication and age and size of the patient. Here, we include basic technical considera-tions, diagnostic justification, special motion considerations,

Table 2 Typical motion management strategies and fields-of-view(FOV) for various age categories in pediatric liver imaging

|Age|Motion management|Axial FOV (mm)|Coronal FOV (mm)| |---|---|---|---| |Infant/neonate(<6 months)|Free-breathing. Feed and bundle up to 6 months|200|220| |Toddler(6 months-6 years)|Free-breathing. Usually sedated|250|280| |School-aged(>6 years)|Non-sedated free-breathing exams are possible in some children|300|320| |Tenager|Breath-hold coaching with motion mitigation|330|340|

Pediatric Radiology

Table 3 Common sequences for pediatric liver MR

Liver anatomy and general parenchymal evaluation

Axial T2 FS(optional axial T2 non-FS, optional coronal T2)

Axial T1 in/opposed-phase or Dixon(IP/OP/F/W)

Optional: axial 2D bSSFP

Optional: axial T1 FS post-contrast

Additional sequences for liver mass

Axial DWI(B=50,800 s/mm2)

Axial T1 FS pre-contrast

Axial T1 FS dynamic post-contrast(30 s, 60 s, 3 min)

Coronal T1 FS post-contrast(3 min)

Additional sequences for biliary imaging

Coronal 2D thick slab T2 SS-FSE MRCP

Coronal 3D T2 MRCP

Axial/coronal T1 FS hepatobiliary-phase contrast-enhanced MRCP

Additional sequences for quantitative evaluation

Iron quantification: R2*/R2

Fat quantification: PDFF

Stiffness quantification: MR elastography

Additional sequences for vascular imaging

3D spoiled GRE (post-contrast MRA): GBCA

3D spoiled GRE (post-contrast MRA): ferumoxytol

Time-resolved MRA

and other tips for common sequences used in pediatric liver evaluation(Table 3). More detailed technical parameters for consideration are included in the supplemental materials.

Liver anatomy and general parenchymal evaluation

The following sequences are typically included in all liver protocols.

1.Axial T2 FS(optional axial T2 non-FS; optional coronal T2)

(a) Technical considerations:

(i) FOV: The complete abdomen(the lung base to iliac crest) for axial. For optional coronal,include the liver and through as much of the pelvis as possible.

(ii) Use an appropriate TE for respective magnet strength to ensure adequate T2-weighting;between approximately 80-100 ms should be used at 1.5 T and 70-100 ms at 3 T[18].

(iii) Optional: axial T2 non-FS and coronal T2.Use of single-shot and avoidance of fat satu-ration can speed this additional sequence [12].Orthogonal plane can supplement anatomic characterization.

(b) Justification: Provides characterization of paren-chyma, biliary tract, and any mass. Fat satura-tion during T2-weighted liver imaging highlights T2-hyperintense areas of liver parenchymal fibro-sis and ema/demainflammation in surrounding tis-sues, and differentiates between T2 hyperintensity secondary to macroscopic fat versus fluid. Often,vessels demonstrate flow-voids on T2-weighted imaging, which can suggest patency, particularly useful when post-gadoxetate sequences may not fully opacify the vessel.

(c)Motion considerations: Speed of single-shot tech-nique can mitigate motion artifacts, but T2 con-trast resolution can be lower on single-shot than on traditional fast/turbo spin-echo sequences or STIR[12, 13].

(d) Pitfalls and pearls/tips and tricks:

(i) Fat saturation increases sequence duration.

(ii) Pre-contrast STIR can provide more homogeneous fat suppression than fat-sat T2 spin-echo sequences. Note that STIR should not be used post-contrast, such as during"dead time" prior to hepatobiliary phase, as gadolinium T1 effects can alter STIR signal intensity[19].

(iii) Fast/turbo spin-echo or single-shot T2-weighted sequences could be per-formed post-contrast with no substantial effects on liver signal.

(iv)In large, obese pediatric patients and/or patients with ascites, standing wave/die-lectric effect artifacts can worsen at higher magnet strengths, i.e., 3 T; consider use of<3 T MR scanner for these patients.

2.Axial T1 in-phase/opposed-phase or Dixon

(a) Technical considerations:

(i) FOV: The complete abdomen(the lung base to iliac crest)

(b) Justification: Provides anatomic detail, parenchy-mal T1 signal assessment, chemical shift charac-terization of fat/lipid and potentially iron content,and evaluation of any mass.

(c) Motion considerations: Conventional 2D GRE in-/opposed-phase sequence can require multi-ple breath-holds; a saturation band may decrease motion artifact but does increase scan time[20].Newer, single breath-hold and free-breathing 3D Dixon techniques are available, which can pro-vide, in a single sequence, the in-/opposed-phase reconstructions, along with additional water-only and fat-only reconstructions[19].

Pediatric Radiology

(d) Pitfalls and pearls/tips and tricks:

(i) In-phase sequence is typically derived from longer TE than opposed-phase, which results in relative in-phase susceptibility signal drop-out(e.g., from iron, air).

(ii)Selection of in-phase TE shorter than opposed-phase TE should generally be avoided if possi-ble, as relative opposed-phase signal dropout can confusingly be from either lipid or susceptibility in this scenario.

(iii) Fat-water swapping artifacts can occur with the Dixon technique, which can impact the whole image or regions of the image. When present, all four Dixon reconstructions should be sent for evaluation. Alternatively, all four reconstructions can be sent routinely.

3. Axial bSSFP(optional)

(a) Technical considerations:

(i) FOV: The complete abdomen(the lung base to iliac crest) or entire biliary tree

(ii) Field strength: Greater susceptibility artifact at 3 T

(b) Justification: Balanced steady-state free preces-sion sequence(using the shortest possible TR to decrease susceptibility) can be a supplemen-tal, relatively motion-insensitive sequence for ductal and vascular characterization due to bright blood appearance, when motion limits other T2 sequences[12]. Radiofrequency pulses maintain a steady state of longitudinal and transverse mag-netization, which results in a mixed contrast image based on T2/T1 ratios and leads to bright signal in both arteries, veins, and bile ducts.

(c) Motion considerations: Respiratory navigation is usually needed as scan time exceeds ability to breath-hold.

(d) Pitfalls and pearls/tips and tricks:

(i) Balanced steady-state free precession can be employed for both 2D and 3D MRCP. This technique, based on gradient-echo sequences,utilizes short repetition times, resulting in brief acquisition times even less than a second per slice. While typically performed with a breath-hold technique, it can also be executed using a respiratory-gated approach.

(ii) For vascular imaging, coronal 3D isotropic acquisition with multiplanar reformats yields the shortest scan time and T2-preparation pulses and fat saturation can be added to increase vessel conspicuity.

(iii) SSFP is very sensitive to field inhomogeneities which can lead to extensive banding artifacts

if post-surgical material is present and/or at air interfaces(such as bowel gas and lung bases when considering liver MRA).

(iv) Flow artifact in hepatic and portal veins as well as tortuous collaterals/varices can occasionally limit assessment of the portal and systemic venous structures.

4. T1 FS post-contrast(optional): Note that post-contrast T1 FS imaging for liver MR typically entails dynamic,multi- and/or hepatobiliary phases of contrast, to maxi-mize benefit, as detailed in indicated protocols below.A single venous phase, post gadolinium-based contrast T1 FS sequence can be pursued for general abdominal evaluation(e.g., extrahepatic neuroblastoma) but typi-cally is insufficient for specific pediatric liver MR indi-cations.

Liver mass

The American College of Radiology(ACR) Pediatric Liver Imaging Reporting and Data Systems(LI-RADS) Work-group has published consensus imaging recommendations and guidance for pediatric liver tumors, with a special focus on hepatoblastoma and hepatocellular carcinoma, that sup-port the PRETEXT system[3]. Similar consensus imag-ing recommendations, built and supported by members of the Children's Oncology Group and Society for Pediatric Radiology, have been published as well[21]. Within these recommendations, MRI has a prominent role in evaluating pediatric liver masses owing to its superior soft tissue con-trast and dynamic vascular imaging capabilities.

1. Axial DWI

a Technical considerations:

i. FOV: Complete abdomen(lung base to iliac crest)

b Justification: DWI/ADC reveals areas of tumor with increased cellularity and along with assessment on post-contrast imaging can help to identify areas to target if biopsy is performed. DWI in correlation with post-contrast imaging can also help identify multifocal disease within the liver, a finding of high significance when the mass is suspected to be a pri-mary malignant tumor given the potential impact on PRETEXT and subsequent treatment decisions.

c Motion considerations: Prone to artifacts from motion and magnetic field inhomogeneity. Naviga-tor triggering offers superior motion compensation,resulting in higher-quality images but at the expense of longer scan times. Signal averaging is more time-efficient but is susceptible to increased motion arti-facts. Newer simultaneous multi-slice(SMS) tech-

Pediatric Radiology

niques provide substantial time savings and can be particularly helpful for DWI imaging [13,22,23].

d Pitfalls and Pearls/Tips and Tricks: b-values of 50-100 s/mm ${}^{2}$ and 800 s/mm ${}^{2}$ generally perform well for liver mass evaluation. Lower b-values func-tion to reveal most masses and mass components that have T2 hyperintensity while the higher B value in conjunction with ADC serves to elucidate those tumors with elements that restrict diffusion [24].DWI can be performed prior to or following con-trast injection; can fill time while waiting for the 20-minute hepatobiliary phase.

2. Axial T1 FS 3D GRE pre-contrast

a Technical considerations:

i. FOV: Complete abdomen(lung base to iliac crest)

b Justification: Serves as a tool for assessment of the intrinsic T1 intensity of liver lesions, including the presence of intralesional hemorrhage, as well as a reference comparison for lesion enhancement on dynamic post-contrast imaging. This sequence may be optional when pre-contrast imaging includes Dixon imaging of the abdomen, generating a water-only image.

c Motion considerations: Breath-hold and free-breath-ing sequence options are available. Breath-hold options are generally shorter sequences and may per-form better in children who are able to comply with breathing instructions. Additional motion considera-tions addressed in dynamic post-contrast section.

d Pitfalls and Pearls/Tips and Tricks: Pre-contrast T1 weighted sequences can be useful to identify liver lesion boundaries and interfaces with liver vascula-ture, helping to determine anatomic position within the liver. This sequence may also be useful for iden-tifying multifocal disease when liver masses have signal intensity either above or below background liver parenchyma. If the breath hold is estimated to be just a little too long for the patient, the field of view can be reduced to cover the liver and kidneys.

3.Axial dynamic post-contrast

a Overall technical considerations: Generally a 3D GRE or Dixon sequence is used.

i. FOV: Complete abdomen(lung base to iliac crest)

<BeginOfPara>ii. Contrast: Both extracellular and hepatobiliary contrast agents can be used to evaluate liver masses in children, with the choice hinging on the suspected pathology of the mass based on prior imaging, clinical, and laboratory findings.Of the most common liver tumors encountered

<EndOfPara>in children, vascular tumors are generally best evaluated with extracellular agents[25] while primary and secondary malignant liver tumors in all ages and focal nodular hyperplasia in old-er children and adolescents are generally bet-ter evaluated with hepatobiliary agents due to the retention of the hepatobiliary agent within functioning hepatocytes[3, 26, 27]. Two options for hepatobiliary contrast agents are available,gadoxetate disodium(Gd-EOB-DTPA, Eovist/Primovist; Bayer, Leverkusen, Germany) and gadobenate dimeglumine(Gd-BOPTA, Multi-Hance; Bracco Diagnostic Inc., Milan, Italy).Gadoxetate is generally preferred due to the provision of hepatobiliary phase imaging at 20 minutes compared to the 45-60 minutes needed for imaging with gadobenate. Many pediatric imaging centers administer a "double dose"(0.05 mmol/kg rather than the FDA-approved dose of 0.025 mmol/kg) of gadoxetate disodium for improved vascular imaging [3]

b Motion considerations: Some patients who are not under anesthesia during their MRI develop transient dyspnea after intravenous injection of gadoxate that may lead to respiratory motion resulting in degradation of image quality, especially during the arterial phase of imaging [28, 29]. Free-breathing,radial 3D GRE acquisitions (e.g., GRASP) can help improve image quality by compensating for respira-tory motion. When conventional 3D GRE sequences(e.g., VIBE) are used in concert with breath-hold technique, the NSA can be increased (e.g., NSA 2) to help compensate for respiratory motion-related artifacts.

c Pitfalls and Pearls/Tips and Tricks:

i. Contrast timing - When using conventional 3D GRE sequences, image acquisition timing is commonly based on a standardized delay meth-odology where imaging for the arterial, portal venous, etc. phases commences after an ap-propriate time interval from the initiation of IV contrast material injection (see contrast phase timing below and Table 4). Another option is to use fluororo-triggering with the descending thoracic aorta used as a monitoring region of interest with imaging initiation triggered con-trast is first visualized in the descending aorta,followed by preset delays between arterial and subsequent phases. Alternatively, sequences using free-breathing radial 3D techniques may initiate imaging prior to contrast injection and have continuous imaging after the initiation of IV contrast material administration for a defined

Pediatric Radiology

Table 4 Contrast timing for liver mass evaluation

|Hepatobiliary contrast for liver mass|||| |---|---|---|---| |Mandatory sequences|Mandatory sequences|Performance notes|Utility| |Dynamic post-contrast|||| |Axial T1 FS 3D GRE-Late Arterial|16-20 s after injection|Assessment of arterial phase enhancement characteristics; assess- ment of hepatic artery. Note: For late arterial phase contrast should be present in the portal venous system, but not the hepatic veins|| |Axial T1 FS 3D GRE-Portal Venous|45-60 s after injection|Assessment of portal venous phase enhancement characteristics; assessment of portal vein|| |Axial T1 FS 3D GRE-Transitional|2-5 min after injection|Assessment of transitional phase enhancement characteristics; assessment of hepatic veins|| |Axial T1 FS 3D GRE-Hepatobiliary|15-20 min (gadoxate) 45-60 min (gadobenate)|Assessment of hepatobiliary phase enhancement characteristics and retention or lack thereof of the contrast agent; assessment of the biliary tree|| |Coronal T1 FS 3D GRE-Hepatobiliary|Optional; 15-20 min|Assessment of hepatobiliary phase enhancement characteristics and retention or lack thereof of the contrast agent; assessment of the biliary tree|| |Extracellular contrast for liver mass|16-20 s after injection|Assessment of arterial phase enhancement characteristics; assess- ment of hepatic artery. Note: For late arterial phase contrast should be present in the portal venous system, but not the hepatic veins|| |Axial T1 FS 3D GRE-Late Arterial|16-20 s after injection|Assessment of portal venous phase enhancement characteristics; assessment of portal vein|| |Axial T1 FS 3D GRE-Portal Venous|45-60 s after injection|Assessment of portal venous phase enhancement characteristics; assessment of hepatic veins|| |Axial T1 FS 3D GRE-Delayed|2-5 min after injection|Assessment of delayed phase enhancement characteristics; assess- ment of hepatic veins|| |Coronal T1 F3 DS GRE-Delayed|Optional; 2-5 min after injection|Assessment of delayed phase enhancement characteristics; assess- ment of hepatic veins||

length of time, allowing the user to set pre-de-termined dynamic imaging phases that are ren-dered available for clinical review.

<BeginOfPara>ii. Gadoxetate- Hepatic vascular assessment, an im-portant part of the assessment when evaluating liver tumors, may be limited by the smaller dose of contrast with gadoxetate, though doubling the dose from 0.025 mmol/kg(0.1 ml/kg) to 0.05 mm/kg(0.2 ml/kg) and infusing at a slower rate can improve vascular enhancement[30,31].

d Axial T1 FS-Late Arterial Phase

i. Timing:16-20 seconds after contrast material injection

<BeginOfPara>ii. Justification: Assessment of lesion arterial phase enhancement characteristics; assessment of the hepatic arteries.

e.Axial T1 FS-Portal Venous Phase

i. Timing:45-60 seconds after contrast material injection

ii. Justification: Assessment of lesion portal venous phase enhancement characteristics; assessment of the portal veins.

f Axial T1 FS-Delayed/Transitional Phase

i. Timing:2-5 minutes after contrast material in-jection

Ⅱ. Justification: Assessment of lesion delayed/transitional phase enhancement characteristics;assessment of hepatic veins.

g Axial T1 FS-Hepatobiliary Phase

i. Timing:20 minutes (gadoxetate),45-60 minutes (gadobenate)

ii. Technique: 3D GRE vs. stack of stars(StarVIBE - Siemens, LAVA Star-GE Health-care, 3D Vane XD - Philips)

iii. Justification: Assessment of lesion hepatobiliary phase enhancement characteristics and retention or lack thereof of the contrast agent; assessment of the biliary tree.

4.Coronal T1 FS-Hepatobiliary (Optional)

i. Timing:15-20 minutes(gadoxetate),45-60 minutes(gadobenate)

ii. Justification: Assessment of lesion hepatobiliary phase enhancement characteristics and retention or lack thereof of the contrast agent; assessment of the biliary tree.

Pediatric Radiology

Biliary disease/MRCP

Magnetic resonance cholangiopancreatography(MRCP) is an MR imaging technique that utilizes heavily T2-weighted sequences to highlight stationary or slow-moving fluid within the biliary tree, typically in the coronal plane. MRCP can be performed using either single-shot breath-hold or respiratory-triggered techniques [32,33]. In recent years,there have been significant advancements in MRCP technol-ogy, including the development of newer pulse sequences and imaging protocols, which have improved image quality,reduced acquisition time, and expanded the range of indi-cations for MRCP [34]. The application of deep learning techniques can further increase image quality [35].

1. Traditional 2D thick slab T2 SS-FSE

a Technical considerations:

i. FOV: Entire biliary tree

ii. Coronal oblique acquisitions parallel to the com-mon bile duct

b Justification: Can be used to quickly capture clear and comprehensive overview images of the biliary tree as the thick slices allow visualization of large portions of the biliary tree within each slab.

<BeginOfPara>d Motion considerations: Typically requires breath holds so prone to motion artifact.

d Pitfalls and Pearls/Tips and Tricks: Long readout period can result in blurring in the phase-encoding direction [36]

2. 3D-FSE MRCP

a Technical considerations:

i. Typically isotropic 3D-fast spin echo (SPACE(Siemens Healthcare), Cube (GE Healthcare),and 3D GRASE or VISTA(Philips Healthcare))

Ⅱ. FOV: Entire biliary tree

iii. Field strength: 1.5 or 3T. Although 3T provides higher signal and smaller voxels, it is more like-ly to have dielectric effect artifact and suscepti-bility in the case of biliary stents [37, 38].

b Justification: Compared to 3D-TSE, 3D-FSE sequences are faster and incur lower SAR. 3D sequences utilizing compressed sensing can decrease imaging time to allow for single breath-hold isovoxel acquisitions while decreasing artifact from respira-tory motion and improving visualization of the bile and pancreatic ducts [39-41]. Isotropic, high-resolu-tion images can be reformatted into different planes.

c Motion considerations:

i. Prone to motion artifact

d Pitfalls and Pearls/Tips and Tricks: Respiratory trig-gering can prolong imaging time, but this can be mitigated with compressed sensing. It may be advan-

tageous to obtain 3D MRCP acquisitions at the begin-ning of the examinations for pediatric patients when patients may be more compliant and less fatigued.

3. Hepatobiliary contrast agent-enhanced MRCP (with gadoxate disodium, optional)

a Technical considerations:

i. FOV: Entire biliary tree

ii. Typically use 3D T1 GRE sequences with fat suppression

iii. Can be acquired in axial and/or coronal planes

iv. Timing:15-20 minutes following contrast in-jection

b Justification: Beyond T2-weighted imaging tech-niques, T1-weighted sequences with hepatobiliary-specific contrast agents can be used for MRCP imag-ing to enhance visualization of the bile ducts and improve diagnostic accuracy. Since these contrast agents are excreted via the biliary system, they can allow for excellent visualization of the bile ducts [42].

c Motion considerations: As discussed in the liver mass section above.

d Pitfalls and Pearls/Tips and Tricks:

i. Very useful to evaluate for bile leaks and to as-sess communication with hepatic cystic lesions Ⅱ. The only MRCP sequence that requires contrast administration, which prolongs imaging time

<EndOfPara>iii. Lower image contrast and spatial resolution compared to conventional T2 MRCP

Quantitative parenchymal characterization

Quantitative liver imaging has become routine for pediatric patients in the non-invasive evaluation of hepatic steato-sis, iron overload, and fibrosis. These exams allow patients to avoid invasive biopsies to diagnose and longitudinally monitor fat, iron, and stiffness levels. Upcoming quantita-tive sequences(e.g., native T1, corrected T1, T1 rho, and T2 mapping) are likely to emerge as additional tools for diagnosing liver inflammation or fibrosis. However, for the purposes of these guidelines, we will review basic recom-mendations for performing liver iron and fat quantification and MR elastography in children. Special challenges in chil-dren include the necessity for breath-holds and inaccurate automated liver segmentation tools.

Although there are several options for the quantification of fat and iron, the most common commercially available sequence allows the simultaneous quantification of fat and iron with a single breath-hold acquisition with coverage of the entire liver. Confounder-corrected chemical shift-encoded(CSE) multi-echo spoiled gradient recalled echo(mSGRE)enables whole-liver estimation of proton density fat-fraction

Pediatric Radiology

(PDFF) and simultaneous fat- and noise-corrected estimation of R2*(1/T2*), a quantitative biomarker of liver iron concen-tration (LIC) [4]. Thus, listed below are general guidelines only as standardization continues to advance.

1. Iron quantification: R2*/R2

- General: The presence of iron accelerates the decay of transverse magnetization in water protons, caus-ing loss of signal in hepatic parenchyma in T2- and T2*-weighted imaging, which can be quantified by the transverse relaxation rates(T2 or T2*, seconds)or rates(R2 or R2*, 1/s).

a Technical considerations:

i. FOV: Entire liver

ii. Field strength: 1.5T or 3T. Conversion to liver iron content requires different calibration equations based on field strength. 1.5 T may be preferred in patients with known or suspected severe iron overload [43].

b R2* relaxometry

i. Justification: To determine the relaxation rate R2* of the gradient echo signal, multiple gra-dient-echo images are typically acquired with increasing echo times within the same repeti-tion time(TR). Reconstruction of the R2* maps includes a fitting of signal intensity decay to estimate the R2* relaxation rate, which is highly correlated and has a linear relationship with liver iron content(LIC). In 2023, a col-laborative effort by the Society for Abdominal Radiology(SAR) and the European Society for Gastrointestinal and Abdominal Radiology(ESGAR) concluded that when available, con-founder-corrected R2*-based LIC quantification is the most practical method with the strongest level of evidence for accurate and reproducible quantification of LIC[43].

<BeginOfPara>ii. Motion considerations: R2* mapping can be performed using parallel imaging to facilitate whole liver coverage in a single short breath-hold(8-12 seconds) while covering the whole liver volume in a 3D acquisition. Strategies for implementing the above sequences in a child who cannot hold their breath include increas-ing the signal averages and performing while free-breathing, reducing coverage to a single slice(reducing acquisition time to< 5 sec), and using respiratory navigation. New techniques utilizing 3D stack-of-stars radial acquisition[44] or multirepetition flip-angle modulated(FAM) acquisitions in combination with non-local means(NLM)-based motion-corrected averaging[45] show promise as free-breathing

techniques for fat and iron quantification in children.

Ⅲ. Pitfalls and Pearls/Tips and Tricks: In patients with severe iron overload, the hepatic parenchymal sig-nal may decay too rapidly for standard msec GRE sequences. In these patients, shortening the initial TE to less than 1 msec can be helpful.

iv. Post-processing: Conversion of R2* values to LIC requires field strength-specific calibration equations [46]. Although there are emerg-ing automated segmented options, typically,regions of interest (ROIs) are drawn manually on the reconstructed R2* map to obtain R2*values.There is currently no consensus on how to draw ROIs. A common recommendation includes drawing at least 4 circular ROIs with a diameter of 2 cm or greater with at least 2 in the right lobe and at least 1 in the left lobe,avoiding areas of artifact, lesions, and vessels.The average R2* value of the 4 or more ROIs is used to estimate LIC. R2* and LIC values should be reported to the nearest integer [4].

C R2 relaxometry

i. Justification: An alternate relaxometry method for the quantification of iron is a spin-echoac-quisition that uses a 90° radio frequency (RF)pulse followed by a 180° refocusing pulse and then samples the signal. The process is repeated with 5 sets of two-dimensional echo times to sample the signal decay due to the transverse relaxation and calculate the R2 relaxation rate.R2-based LIC quantification is available via the commercial products Ferriscan and Ferrismart(Resonance Health). These products are FDA-approved for LIC quantification and require an additional cost.

Ⅱ. Motion considerations: Images are acquired over 10-20 minutes of free breathing.

iii. Post-processing: For Ferriscan, acquired im-ages are submitted electronically to Resonance Health's data processing center for image processing, and an electronic report is returned with the mean R2 value and corresponding LIC value[47]. For Ferrismart, a cloud-based convolutional neural network trained on the database from FerriScan provides real-time analysis and instant LIC reporting[48].

2. Fat quantification

a Technical considerations:

i. FOV: Entire liver

Ⅱ. Field strength: 1.5T or 3T.PDFF values are in-dependent of field strength.

Pediatric Radiology

b Justification: Chemical shift encoded MRI(CSE-MRI) computationally synthesizes T1-weighted fat-only and water-only images, from which a fat fraction(FF) map can be calculated. Confounder correction(e.g., T1 bias and T2* decay) enables whole-liver estimation of proton density fat-frac-tion(PDFF), a quantitative biomarker of triglycer-ide concentration in the liver, and has been robustly validated across field strengths and vendors.

c Motion considerations: See above R2* section.

d Pitfalls and Pearls/Tips and Tricks: Review the source fat and water-only images to ensure there has not been a fat/water swap. Review the auto-segmentation provided by the vendor to ensure the segmentation is accurate.

e Post-processing: See above R2* section for ROI placement discussion. The mean PDFF is reported.Consider including the range if the fat signal is obvi-ously heterogenous.

3. Elastography

a Technical considerations:

i. FOV: Typically 4 axial slices are obtained in the liver, including the widest portions and exclud-ing the heart in the field of view.

ii. Field strength: 1.5T and 3T. Liver stiffness is independent of field strength.

iii. Driver set-up: The rigid passive driver is posi-tioned over the right hepatic lobe, with the center of the paddle to the right of the nipple line. The driver should not extend to the left of the xiphold process except in the smallest patients. Preferen-tially, the paddle is positioned laterally rather than medially. The goal is to position the driver over the largest portion of the liver. The paddle is fas-tened in place tightly with an elastic band. Driver amplitude is adjusted for patient weight [49].

iv. Other considerations: Patients are required to fast before MRE, typically 4-6 hours.

b Justification: The common downstream pathology of multiple pediatric liver diseases(e.g., autoimmune,metabolic, and nonalcoholic fatty liver disease) is the development of liver fibrosis and ultimately cir-rhosis and end-stage liver disease. MR elastography is a technique that allows the non-invasive assess-ment of liver stiffness, a surrogate for hepatic fibro-sis[4]. This technique has a high rate of technical success in children[49].

c Motion considerations: Generally, acquisitions are performed with the patient at end-expiration, which ensures reproducible positioning of the diaphragm and liver[50]. However, end-inspiration breath

holds can be easier for younger children and are an option. Free-breathing options are emerging[51],but not available for routine clinical use today.

d Pitfalls and Pearls/Tips and Tricks: It is important that the paddle is tightly fastened, especially in larger patients, to ensure waves penetrate the liver.In the presence of liver iron there is a higher tech-nical failure rate using 2D GRE-MRE compared to spin echo-echo planar imaging(SE-EPI)[49, 52].MRE can be performed prior to or following con-trast administration with no effect on liver stiffness values. For patient comfort, MRE can performed early in the exam, and if included in a full contrast liver protocol, the paddle can be removed follow-ing MRE for patient comfort for the remainder of the exam.

e Post-processing: ROI measurements should be made by using a freehand ROI tool, sampling the largest portion of the liver on each elastogram while avoid-ing the peripheral 1 cm of the liver, large vessels,extrahepatic tissues, fissures, masses, and the gall-bladder fossa. The arithmetic mean liver stiffness measurement is recorded in kilopascals and reported to the nearest decimal in addition to the range[4].

Liver angiography

At times, dedicated MR angiography(MRA) might be required as part of the liver imaging protocol. Indications can be divided into two main categories:(1) congenital/portosystemic or surgical shunt evaluation and(2) vascu-lar mapping prior to surgery. Vascular mapping could be related to the liver in general in the case of transplant and/or tumor supply/invasion in the case of resection. Inclusion of an MRA depends on institutional practices. Computed tomography angiography(CTA) with pediatric-specific protocols for dose and contrast optimization is a reasonable alternative depending on institutional experience and patient factors[53]. We focus on post-contrast MRA techniques due to their broader clinical use and more reliable vascular visu-alization, whereas non-contrast methods such as balanced steady-state free precession(bSSFP) and inflow-sensitive inversion recovery(IFIR) are limited by flow dependence,motion sensitivity, longer scan times, and challenges in spa-tial resolution, particularly in pediatric patients with altered hemodynamics [54].

1. 3D Spoiled gradient echo(post-contrast 3D MRA):GBCA

a Technical considerations:

i. Field of view and matrix: Coronal 3D isotropic acquisition yields the shortest scan time with best coverage for reformatting in sagittal and

Pediatric Radiology

axial planes

ii. Scan parameters[55]:

1. Keep TE and TR as short as possible (TE 1-2 ms, TR 306 ms)

2. Flip angle optimization(25-45 degrees for best signal to noise ratio(SNR))

b Justification: Contrast-enhanced MRA is a spoiled gradient echo sequence that uses the T1 shortening properties of GBCA and flow-related vessel enhance-ment present at baseline to image the vasculature during and shortly after GBCA injection. Spoiler gradients accentuate T1-W and suppress background signal with greater flip angles providing greater sup-pression of background signal, maximizing vessel contrast. It can be acquired as a short, breath held sequence for pure arterial timing versus a longer acquisition for more detailed vessel assessment that includes both arterial and venous enhancement.

c Motion considerations:

i. Breath hold: If the patient is under general an-esthesia, the team needs to be aware of breath holds prior to starting.

ii. Respiratory-navigated: The target is an accept-ance window of greater than 50%. Quiet free breathing is desired with shallow inspiration and prolonged expiration, to decrease diaphragm variation and prevent hypervilating. Shorter scan times mean less bulk gross motion which may outweigh disadvantages of slight diaphrag-matic blurring on angiographic imaging. Navi-gator placement on the spleen can be employed when imaging the liver. Pleural effusions/con-solidation, asplenia or polysplenia, and iron overload can limit gating on the spleen.

d Contrast considerations:

i. Volume: Contrast volumes are generally low in MRI, which can be a challenge in small children and infants. To increase volume, contrast can be diluted with saline. However, it should be well mixed just before injection and typically should not be diluted by more than 50%. A double dose could be considered with attention to lifetime dose accumulation. The saline flush is impor-tant to ensure the contrast is pushed through the tubing into the patient. In general, the flush can be 10-20 ml or 2 ml/kg in patients< 10 kg.Remember, most tubing holds around 5 mL;therefore, at least 5 mL flush should be given to clear contrast from the tubing.

<BeginOfPara>ii. Rate: A rate of 1 ml/s is usually sufficient for liv-er angiography. Slower rates extend the contrast plateau but decrease concentration of the bolus.

If the injection is too fast, the bolus might be missed. In addition, the high intravascular signal peak of a compact bolus can cause truncation(ringing) artifact, which can be confused for a filling defect. If possible, the technologist can adjust injection rate so the bolus lasts approxi-mately half of the acquisition time. Split bo-lus injection is another approach, which injects half to two thirds of the contrast as a relatively fast bolus and then continues with a very slow infusion(0.5 ml/s or less) of the remainder to preserve some circulating intravascular signal throughout k-space filling.

iii. Timing:Timing:Timing of the bolus depends on injec-tion, patient hemodynamics, and sequence pa-rameters.Time to acquisition can be determined with a test bolus, bolus tracking, or time-resolved imaging. A test bolus is rarely used in pediat-rics because of limited contrast volume in such small patients.In bolus tracking, the full bolus is administered, sampling the artery of interest rapidly using a 2D GRE sequence that begins as contrast is injected. When the technologist sees an increase in signal intensity in the vessel of in-terest, they trigger the diagnostic scan when the signal threshold is reached. Because the scanner is actively scanning, it can be difficult for the patient to hear breath hold commands upon scan triggering. Steady state 3D MRA could follow a time-resolved MRA with no trigger required.

e Pitfalls and Pearls/Tips and Tricks:

i.Tortuous collaterals/varices can be hard to fill,which could lead to misinterpretation of throm-bosis. In steady state imaging, there is loss of separation of arterial and venous phase. This could mask arterial pathology, such as splenic aneurysms, in patients with tortuous varices[56].

ii. Generating subtraction images, maximum in-tensity projections(MIP), and utilizing fat sup-pression and T2-preparation pulses can increase conspicuity of contrast[56].

iii. Subtraction images are often used wherein a precontrast sequence is acquired and then sub-tracted from post-contrast images. If there is a question of narrowing or occlusion on subtrac-tion images, the source images should be refer-enced as artifact can occur from discrepancies between pre- and post-contrast images[56].

iv. In choosing between the conventional, breath-hold triggered CE-MRA and a steady-state 3D navi-gated CE-MRA, the conventional method should be chosen when pure arterial imaging is required

Pediatric Radiology

without venous contamination, the patient cannot cooperate for the time needed for navigated MRA,or there is multi-station coverage required in as-sessment(e.g. bolus chasing)[56].

2. 3D Spoiled gradient echo (post-contrast 3D MRA):Ferumoxytol

a Technical considerations:

i. Scan parameters:

1. Flip angle decreased in comparison with GBCA 3D spoiled gradient echo

2. Differences between 1.5T and 3T: Increase time to inversion and further decrease flip angle at 3T

ii. Safety considerations: The FDA has issued a black box warning for Feraheme®(AMAG Pharmaceuticals, Waltham, Massachusetts) for potential acute severe hypersensitivity, identify-ing rapid infusion of undiluted ferumoxytol as a potential risk factor and recommending infusion over 15 minutes and hemodynamic monitoring for up to 30 minutes after infusion[57]. A multi-center registry, including 311 injections in pedi-atric patients<16 years old, showed that nearly 20 years of ferumoxytol use in MRI appeared to be well tolerated, with a positive safety profile and no severe adverse events[57].

iii. Infusion considerations: The logistics of in-fusion will not be discussed here, but include when(e.g. before or during scanning) and where(e.g. outside or inside scanner) ferumoxytol will be injected[58]. Ferumoxytol intravascular en-hancement can last for hours to days and it can have lasting effects on imaging, including its metabolism by the reticuloendothelial system and T2* effects.

b Justification: Ferumoxytol can be used to achieve superb, long lasting intravascular enhancement. It can be particularly useful in tortuous, dilated ves-sels that might have incomplete mixing or wash-out before uniform opacification is reached with extravascular contrast agents. In addition, the uni-form opacification achieved with ferumoxytol can be useful in outlining filling defects in cases of tumor-thrombus and/or thrombosis in slow-flow, tortuous,and/or compressed vessels

c Motion considerations: Free-breathing with respira-tory navigation is most often used.

d Pitfalls and Pearls/Tips and Tricks: Secondary to the FDA infusion recommendations, the MRA with fer-umoxytol is steady state, meaning venous and arte-rial opacification. Because of the long intravascular half life, if a patient needs a break from scanning,

they can leave the scanner and return when ready.Any GRE T1-weighted sequence can be used after ferumoxytol to assess the vascular enhancement if the MRA specific sequence is failing.

3. Time-resolved MRA

a Technical considerations[59]:

i. FOV: Usually a coronal acquisition

ii. Temporal resolution: Temporal resolution is usu-ally between 1-2 s if arterial evaluation and can be lengthened to 4-6 s if portal or systemic venous.Keyhole fraction is usually between 20-25%.

iii. Spatial resolution: Ideal spatial resolution is be-tween 1.0-1.8 mm^3

iv. Number of dynamic phases: Usually between 8-12 if arterial focus and 10-25 if vascular malformation or venous evaluation.

b Justification: Multiphase imaging allows for sepa-ration of arterial from venous flow and real-time visualization of flow dynamics. Time resolved MRA fills central k-space each acquisition and partially fills periphery each acquisition. The partial acquisi-tions of the periphery are summated to reconstruct a full volume because there should be little to no change if there isn't significant motion. This makes it ideal for dynamic evaluation, such as in shunts and vascular malformations. Lastly, it can provide a maximal amount of arterial contrast with small bolus volumes.

c Motion considerations: Depending on the length of scan needed, breath holding can be used. If being per-formed for venous or slow flow vascular malformation,the scan time usually dictates quiet breathing.

d Pitfalls and Pearls/Tips and Tricks: TR-MRA can be particularly useful when the contrast bolus is small.Attempts can be made to synchronize the respiratory cycle and acquisition; however, encouraging quiet free breathing might be more successful.

Conclusion

These guidelines and suggestions represent a concerted effort from the SPR MRI and Abdominal Imaging Committees to provide expert guidance, suggestions, and rationale for the performance of core elements of pediatric liver MRI. Herein,we have strived to provide the reader with elements of and explanation for what we believe encompasses optimized pediatric liver MRI evaluation. Our hope is this document will serve as a tool for those charged with helping to care for pediatric patients with a wide variety of liver diseases where MRI may provide valuable contributions to care.

Pediatric Radiology

Supplementary Information The online version contains sup-plementary material available at https://doi.org/10.1007/s00247-025-06383-3.

Acknowledgements We gratefully acknowledge Rachel Young, MRI technologist at Cincinnati Children's Hospital, for her expert review and insightful feedback. We also extend our sincere appreciation to the members of the Society for Pediatric Radiology(SPR) MRI and Abdominal Imaging Committees for their thorough review, construc-tive input, and endorsement of the final document.

Author contribution C.E.M.and G.R.S.conceptualized the manuscript and coordinated contributions from the SPR MRI and Abdominal Imaging Committees.J.Y.C.,E.B.M.,and S.H.R.prepared tables.All authors contributed to writing sections of the manuscript, reviewed the content for accuracy and clarity, and approved the final version. The final document was endorsed by the SPR MRI and Abdominal Imaging Committees and approved by the SPR Board of Directors.

Data availability No datasets were generated or analysed during the current study.

Declarations

Conflicts of interest None.

Open Access This article is licensed under a Creative Commons Atri-bution 4.0 International License, which permits use, sharing, adapta-tion, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source,provide a link to the Creative Commons licence, and indicate if changes were made.The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

References

1. Simmons CL, Harper LK, Patel MC et al(2024) Biliary disor-ders, anomalies, and malignancies in children. Radiographics 44:e230109

2. Albers BK, Khanna G(2019) Vascular anomalies of the pediatric liver. Radiographics 39:842-856

3.Schooler GR, Squires JH, Alazraki A et al(2020) Pediatric hepate-toblasta, hepatocellular carcinoma, and other hepatic neo-plasms: consensus imaging recommendations from American College of Radiology Pediatric Liver Reporting and Data System(LI-RADS) Working Group. Radiology 296:493-497

4.Guglielmo FF, Barr RG, Yokoo T et al(2023) Liver fibrosis,fat, and iron evaluation with MRI and fibrosis and fat evalua-tion with US: a practical guide for radiologists. Radiographics 43:e220181

5.Courtier J,Rao AG,Aupinndi SA(2017) Advanced imaging tech-niques in pediatric body MRI. Pediatr Radiol 47:522-533

6.Ahmad R, Hu HH, Krishnamurthy R, Krishnamurthy R(2018)Reducing sedation for pediatric body MRI using accelerated and abbreviated imaging protocols. Pediatr Radiol 48:37-49

7. Darge R, Anupindi SA, Jaramillo D(2011) MR imaging of the abdomen and pelvis in infants, children, and adolescents. Radiol-ogy 261:12-29

8. Nardone B, Saddleton E, Laumann AE et al(2014) Pediatric nephrogenic systemic fibrosis is rarely reported: a RADAR report.

9.Stanescu AL, Shaw DW, Murata N et al(2020) Brain tissue goado-linium retention in pediatric patients after contrast-enhanced mag-netic resonance examples: pathological confirmation. Pediatr Radiol 50:388-396

10.Chernyak V, Kim J, Rozenblit AM et al(2011) Hepatic enhance-ment during the hepatobiliary phase after goadetate disodium administration in patients with chronic liver disease: the role of laboratory factors.J Magn Reson Imaging 34:301-309

11. Dong S-Z, Zhu M, Bulas D(2019) Techniques for minimizing sedimentation in pediatric MRI.J Magn Reson Imaging 50:1047-1054

12.Chavhan GB,Schooler GR, Tang ER et al(2023) Optimizing imaging of pediatric liver lesions: guidelines from the Pediatric LI-RADS working group. Radiographics 43:e220043

13. Gallo-Bernal S, Bedoya MA, Gee MS, Jaimes C(2022) Pediat-ric magnetic resonance imaging: faster is better. Pediatr Radiol.https://doi.org/10.1007/s00247-022-05529-x

14. Kozak BM, Jaimes C, Kirsch J, Gee MS(2020) MRI tech-niques to decrease imaging times in children. Radiographics 40:485-502

15. Feinberg DA, Hale JD, Watts JC et al(1986) Halving MR imag-ing time by conjugation: demonstration at 3.5 kG. Radiology 161:527-531

16. Zhang T, Chowdhury S, Lustig M et al(2014) Clinical perfor-mance of contrast enhanced abdominal pediatric MRI with fast combined parallel imaging compressed sensing reconstruction: fast CS in abdominal pediatric MRI. J Magn Reson Imaging 40:13-25

17.Chung T, Dillman JR.(2023) Deep learning image reconstruction:a tremendous advance for clinical MRI but be careful.... Pediatr Radiol 53:2157-2158

18.Yeh B, Chair, Martin DR, et al(2017) ACR-SAR-SPR practice parameter for the performance of magnetic resonance imaging(MRI) of the liver

19. Shetty AS, Fraum TJ, Ludwig DR et al(2022) Body MRI: imag-ing protocols, techniques, and lessons learned. Radiographics 42:2054-2074

20. Serai SD, Hu HH, Ahmad R et al(2020) Newly developed meth-ods for reducing motion artifacts in pediatric abdominal MRI: tips and pearls. AJ R Am J Roentgenol 214:1042-1053

21.Schooler GR, Infante JC, Acord M et al(2022) Imaging of pedi-atric liver tumors: a COG diagnostic imaging committee/SPR oncology committee white paper. Pediatr Blood Cancer. https://doi.org/10.1002/pbc.29965

22.Jerome NP,Orton MR,d'Arcy JA et al(2015) Use of the temporal median and trimmed mean mitigates effects of respiratory motion in multiple-acquisition abdominal diffusion imaging. Phys Med Biol 60:N9-20

23. Taron J, Martirosian P, Erb M et al(2016) Simultaneous multislice diffusion-weighted MRI of the liver: analysis of different breath-ing schemes in comparison to standard sequences. J Magn Reson Imaging 44:865-879

24. Grasparil ADI, Gupta H, Sheybani E, Chavhan GB(2019) Low b-value(50-100) diffusion-weighted images detect significantly more hyperintense liver lesions in children than T2-weighted images. Pediatr Radiol 49:1299-1305

25.Chavhan GB, Farras Roca L, Coblentz AC(2022) Liver magnetic resonance imaging: how we do it. Pediatr Radiol 52:167-176

26.Meyers AB,Towbin AJ,Geller JI, Podberesky DJ(2012) Hepato-blastoma imaging with gadoetate disodium-enhanced MRI-typi-cal, atypical, pre- and post-treatment evaluation. Pediatr Radiol 42:859-866

27. Kolbe AB, Podberesky DJ, Zhang B, Towbin AJ(2015) The impact of hepatocyte phase imaging from infancy to young

Pediatric Radiology

adulthood in patients with a known or suspected liver lesion.Pediatr Radiol 45:354-365

28. Gilligan LA, Trout AT, Anton CG et al(2019) Respiratory motion in children and young adults undergoing liver magnetic resonance imaging with intravenous goxatete disodium contrast material.Pediatr Radiol 49:1171-1176

29. Lanier H, Wallace A, Khanna G(2020) Rate of gadoxateo disco-ium(Eovist®) induced transient respiratory motion in children and young adults. Abdom Radiol(NY) 45:101-106

30. Chavhan GB,Shelmerdine S, Jhaveri K, Babyn PS(2016) Liver MR imaging in children: current concepts and technique. Radio-graphics 36:1517-1532

31.Reimer P, Rummery EJ, Shamsi K et al(1996) Phase II clinical evaluation of Gd-EOB-DTPA: dose, safety aspects, and pulse sequence. Radiology 199:177-183

32. Griffin N, Charles-Edwards G, Grant LA(2012) Magnetic reso-nance cholangiopancreatography: the ABC of MRCP. Insights Imaging 3:11-21

33.Vidal BPC, Lahan-Martins D. Penachim TJ et al(2020) MR cholangiopancreatography: what every radiology resident must know. Radiographics 40:1263-1264

34. Kraus MS, Coblentz AC, Deshpande VS et al(2022) State-of-the-art magnetic resonance imaging sequences for pedi-atric body imaging. Pediatr Radiol. https://doi.org/10.1007/s00247-022-05528-y

35. Zhang Y, Peng W, Xiao Y et al(2023) Rapid 3D breath-hold MR cholangiopancreatography using deep learning-constrained compressed sensing reconstruction. Eur Radiol 33:2500-2509

36. Welle CL, Miller FH, Yeh BM(2020) Advances in MR imag-ing of the biliary tract. Magn Reson Imaging Clin N Am 28:341-352

37. Almehdar A, Chavhan GB(2013) Mr cholangiopancreatography at 3.0 T in children: diagnostic quality and ability in assessment of common paediatric pancreatobiliary pathology. Br J Radiol 86:20130036

38. Glenn, Trout AT, Kocaoglu M et al(2022) Patient- and examination-related predictors of 3D MRCP image quality in children.AJR Am J Roentgenol 218:910-916

39. Yoon JH, Lee SM, Kang H-J et al(2017) Clinical feasibility of 3-dimensional magnetic resonance cholangiopancreatography using compressed sensing: comparison of image quality and diagnostic performance. Invest Radiol 52:612

40. Chevallier O, Escande H, Ambarki K et al(2021) Single-breath-hold MRI-SPACE cholangiopreatography with compressed sensing versus conventional respiratory-triggered MRI-SPACE cholangiopreatography at 3 Tesla: comparison of image quality and diagnostic confidence. Diagnostics 11:1886

41. Sodhi KS, Maralakute M, Bhatia A et al(2023) Utility of the new faster compressed SENSE MRCP at 3 Tesla MRI in chil-dren with pancreatitis. Indian J Pediatr 90:1210-1215

42. Shet NS, Flynn JF, Maloney E,Iyer RS(2020) Use of Eovist in pediatric patients: pearls and pitfalls. Curr Probl Diagn Radiol 49:266-274

43. Reeder SB, Yokoo T, Franca M et al(2023) Quantification of liver iron overload with MRI: review and guidelines from the ESGAR and SAR. Radiology 307:e221856

44. Rohani SC, Morin CE, Zhong X et al(2021) Hepatic iron quantification using a free-breathing 3D radial gradient echo

technique and validation with a 2D biopsy-calibrated R2* relax-ometry method. J Magn Reson Imaging 55:1407-1416

45. Starekova J, Zhao R, Colgan TJ et al(2022) Improved free-breathing liver fat and iron quantification using a 2D chemical shift-encoded MRI with flip angle modulation and motion-cor-rected averaging. Eur Radiol 32:5458-5467

46. Herrando D, Zhao R, Yuan Q, et al(2022) Multicenter repro-ducibility of liver in quantification with 1.5-T and 3.0-T MRI.Radiology 213256

47. Pavitt HL, Ayokin Y,El-Beshlawy A et al(2011) The effect of reducing repetition time TR on the measurement of liver R2 for the purpose of measuring liver iron concentration. Magn Reson Med 65:1346-1351

48.(2019) FerriSmart- automated MRI measurement of liver iron concentration. https://ferriscan.com/ferrismart/. Accessed 4 May 2021

49.Joshi M, Dillman JR, Towbin AJ et al(2017) MR elastogra-phy: high rate of technical success in pediatric and young adult patients. Pediatr Radiol 47:838-843

50. Depin K M, Welle CL, GugliemoF FF et al(2022) Magnetic reso-nance elastography of the liver: everything you need to know to get started. Abdom Radiol(NY) 47:94-114

51.Morin CE, Dillman JR, Serai SD et al(2018) Comparison of standard breath-held, free-breathing, and compressed sensing 2D gradient-recalled echo MR elastography techniques for evaluating liver stiffness in JR Rometgenol 211:W279-W287

52. Serai SD, Trout AT(2019) Can MR elastography be used to meas-ure liver stiffness in patients with iron overload? Abdom Radiol(NY) 44:104-109

53.Verhagen MV,Dikkers R, de Kleine RH et al(2021) Assessment of hepatic artery anatomy in pediatric liver transplant recipients:MR angiography versus CT angiography. Pediatr Transplant 25:e14002

54. Fleccs JB, Artz NS, Mitchell GS, Chan SS(2022) Non-contrast magnetic resonance angiography/venography techniques: what are my options? Pediatr Radiol 52:271-284

55.Krishnamurthy R, Malone L, Lyons K et al(2016) Body MR angiography in children: how we do it. Pediatr Radiol 46:748-763

56.Liszewski MC,Kurian J(2020) Tailored optimization of pediatric body MR angiography for successful outcomes in thoracic appli-cations.AJR Am J Roentgenol 214:1031-1041

57. Nguyen K-L, Yoshida T, Kathuria-Prakash N et al(2019) Multi-center safety and practice for off-label diagnostic use of ferumoxy-tol in MRI. Radiology 293:554-564

58.Shah R, VanSyckel A, Popescu AR et al(2023) Guide to use of ferumoxytol for hepatic vascular assessment as part of dual contrast MRI. Pediatr Radiol. https://doi.org/10.1007/s00247-023-05737-z

59. Krishnamurthy R, Bahouth SM, Muthupillai R(2016) 4D con-trast-enhanced MR angiography with the keyhole technique in children: technique and clinical applications. Radiographics 36:523-537

Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Pediatric Radiology

Authors and Affiliations

Cara E.Morin ${ }^{1,2}$ .Joseph Y.Cao ${ }^{3}$.Lindsay M.Griffin ${ }^{4}$ .Erin B.Macdonald ${ }^{3}$ .Morgan P.McBee ${ }^{5}$ .Bruce L.McHam ${ }^{6}$ .Scott H.Robertson ${ }^{3}$ .Elizabeth Tang ${ }^{7}$ . Gary R. Schooler ${ }^{1,2}$

※

■Cara E.Morin

cara.morin@cchmc.org

Joseph Y.Cao

joseph.cao@duke.edu

Lindsay M. Griffin

lindsay.griffin@radpartners.com

Erin B. Macdonald

erin.macdonald@duke.edu

Morgan P. McBee

mcbeem@musc.edu

Bruce L. McHam

mcham.bruce@marshfieldclinic.org

Scott H. Robertson

scott.robertson@duke.edu

Elizabeth Tang

elizabeth.tang@childrenscolorado.org

Gary R.Schooler

gary.schooler@cchmc.org

1 Cincinnati Children's Hospital Medical Center, 3333 Burnet Ave, Cincinnati, OH 45229, United States

2 University of Cincinnati, Cincinnati, United States

3 Duke University, Durham, United States

4 Connecticut Children's Medical Center, Hartford,United States

5 Medical University of South Carolina, Charleston,United States

6 Marshfield Clinic, Marshfield, United States

$\boxed{7}$ Children's Hospital Colorado, Aurora, United States