MRI Pancreas – Generic Standard Protocol

1. Executive Summary

Pancreatic MRI occupies a unique and clinically expanding position in abdominal imaging. Unlike the liver — where CT and MRI compete on approximately equal terms for most indications — the pancreas presents a more nuanced picture. CT remains the gold standard for acute pancreatitis severity assessment, peripancreatic vascular evaluation for surgical staging of pancreatic ductal adenocarcinoma (PDAC), and detection of pancreatic calcifications. MRI, combined with MRCP, is the superior modality for duct morphology, cystic lesion characterisation, parenchymal tissue contrast, and long-term surveillance without cumulative ionising radiation.

The generic adult pancreatic MRI protocol described on this page covers the full diagnostic range of pancreatic indications — from cystic lesion surveillance to solid mass characterisation to parenchymal disease assessment. It is built on the minimum standard endorsed by the PRECEDE Consortium for MRI/MRCP of the pancreas [1], adapted for general clinical practice with the ESGAR and ACR frameworks [2, 3].

The intrinsic advantage of MRI over CT for the pancreas derives from two specific physical capabilities that CT cannot match: (1) T1 pre-contrast signal in normal pancreatic parenchyma — the pancreas has the highest intrinsic T1 signal of any abdominal organ due to its rich aqueous protein content and absence of fibrous stroma, making T1-hypointense lesions conspicuous; (2) MRCP — non-invasive, non-irradiating, high-resolution imaging of the pancreatic duct system that is simply not achievable by CT.

1.1 Core Strengths

Duct morphology via MRCP: the main pancreatic duct (MPD), side branches, and the biliary system are visualised in three dimensions without contrast injection or radiation. Ductal dilation, strictures, filling defects, side-branch IPMN communication, and pancreatic divisum are all primary MRCP indications [4, 5].

Pre-contrast T1 signal as a diagnostic tool: normal pancreatic parenchyma is T1-bright on fat-suppressed sequences due to its protein-rich secretory granule content. Diffuse or focal T1 signal loss is one of the most sensitive early markers of pancreatic pathology — chronic pancreatitis, parenchymal fibrosis, and even subtle PDAC produce characteristic T1 signal changes that are invisible on CT [6, 7].

Cystic lesion characterisation: MRI provides superior soft tissue contrast for depicting septations, mural nodules, solid components, and ductal communication within pancreatic cystic lesions compared with CT. It is the preferred modality for surveillance of pancreatic cysts, IPMN, and indeterminate cystic lesions [4, 5, 8, 9].

No ionising radiation: pancreatic cyst surveillance requires serial imaging over years to decades. MRI enables surveillance without cumulative radiation dose — critically important in younger patients and in hereditary high-risk individuals who begin surveillance at 40–45 years [2, 3].

DWI for lesion detection: diffusion restriction in the pancreas indicates high cellularity (PDAC, solid pseudopapillary tumour, some neuroendocrine tumours) and can detect small solid lesions against the bright pre-contrast T1 background.

Functional information via secretin-MRCP (conditional): intravenous secretin stimulates bicarbonate secretion, temporarily dilating the MPD and exocrine secretory units, enabling functional exocrine assessment — not achievable by any CT technique.

1.2 Intrinsic Limitations of the Generic Protocol

CT remains superior for acute pancreatitis and vascular staging: CECT is the standard for CT severity index calculation, necrotising pancreatitis assessment, and vascular involvement evaluation in PDAC staging. MRI is complementary in the sub-acute and chronic setting; it does not replace CT in the acute resuscitation context.

Small solid lesion detectability: PDAC below 1–1.5 cm is challenging for any cross-sectional modality. MRI and CT perform similarly for PDAC staging overall; MRI may be slightly superior for small lesions due to T1 and DWI sensitivity, but neither modality reliably detects sub-centimetre PDAC outside dedicated screening programmes.

Breath-hold dependence: the pancreas is located in the retroperitoneum, adjacent to the duodenum and the stomach, and undergoes several centimetres of craniocaudal displacement with each respiratory cycle. All high-quality pancreatic sequences require consistent breath-holding. The pancreas tolerates motion far less than the liver because: (1) it is smaller and thinner; (2) the MPD occupies only 2–5 mm of the gland width; (3) subtle parenchymal signal changes require pixel-level consistency to detect. Motion degradation is the most common cause of non-diagnostic pancreatic MRI in clinical practice.

MRCP limitations: MRCP depicts luminal morphology but does not show the pancreatic wall, adjacent vasculature, or parenchymal changes with the resolution of direct endoscopy. A non-dilated MPD (< 3 mm) may not be fully characterised on standard MRCP. Secretin-enhanced MRCP adds functional information but requires specific pharmaceutical availability and patient monitoring protocols.

When dedicated child protocols are required: acute pancreatitis severity assessment; PDAC locoregional staging with vascular mapping; hereditary high-risk PDAC screening (dedicated PRECEDE/CAPS protocol); secretin-MRCP for exocrine function testing; autoimmune pancreatitis with IgG4-related disease workup; post-Whipple assessment; IPMN surveillance programme with standardised reporting.

2. Main Clinical Indications

2.1 Standard Indications

Pancreatic cystic lesion surveillance is the most common indication for pancreatic MRI and one where MRI is definitively the preferred modality over CT. Pancreatic cysts are detected incidentally in approximately 2–3% of abdominal CT and MRI studies and in up to 25% of autopsy series. The primary clinical question is whether a cyst has features suggesting malignant potential (worrisome features or high-risk stigmata per Fukuoka/Sendai/international guidelines). MRI with MRCP provides: cyst size and morphology; septations and wall thickness; mural nodule detection; ductal communication; MPD diameter [4, 5, 9]. The generic protocol is sufficient for initial characterisation and routine surveillance. Serial surveillance with consistent protocol parameters is essential for accurate cyst growth assessment.

Indeterminate pancreatic lesion characterisation after CT: when CT demonstrates a pancreatic lesion that is incompletely characterised — particularly cystic or partially cystic lesions, lesions with uncertain solid components, or incidental lesions of unclear nature — MRI with MRCP provides characterisation that CT cannot match. For solid lesions that are hypovascular or isoattenuating on CT, the T1 pre-contrast signal and DWI characteristics of MRI often clarify the nature.

Suspected PDAC with inconclusive CT: MRI adds value when CT is non-diagnostic for PDAC — particularly for small isoattenuating tumours, parenchymal signal changes suggesting ductal obstruction, and for assessment of biliary and pancreatic ductal changes associated with early-stage disease [6, 7]. The generic protocol includes the sequences required to assess the key MRI signs of PDAC (T1 signal loss, duct obstruction, restricted diffusion, delayed enhancement). For formal surgical staging, CT with pancreatic protocol thin slices remains the standard vascular mapping tool.

Chronic pancreatitis assessment: MRI with MRCP provides parenchymal atrophy characterisation, T1 signal changes of fibrosis, ductal morphology (irregularity, strictures, dilated side branches), intraductal stones (T2 dark filling defects), and pseudocyst morphology. Secretin-enhanced MRCP adds functional exocrine assessment but requires specific infrastructure. The generic protocol is adequate for diagnostic characterisation; the secretin-enhanced protocol is a conditional addition when functional assessment is specifically required.

Post-ERCP or post-surgical evaluation: following ERCP, endoscopic sphincterotomy, or pancreatic surgery (particularly partial pancreatectomy or Whipple procedure), MRI with MRCP provides non-invasive assessment of ductal anatomy, anastomotic integrity, pancreatic fistula, and post-surgical collections. The generic protocol requires modification for post-Whipple anatomy (dedicated child protocol).

Autoimmune pancreatitis: the diffuse sausage-shaped pancreatic enlargement with T1 signal loss, hypointense halo, and post-contrast "delayed" enhancement pattern is well characterised on MRI. IgG4-related disease workup may require additional sequences. The generic protocol identifies the typical features; follow-up after steroid treatment is a primary MRI surveillance indication.

Hereditary high-risk individual screening: in individuals with BRCA1/2, PALB2, ATM, Lynch syndrome, familial pancreatic cancer, and Peutz-Jeghers syndrome, annual MRI/MRCP screening is recommended from age 40–50 years by the CAPS consortium and is formally endorsed by the PRECEDE Consortium minimum protocol [1, 2, 3]. The generic protocol closely matches the PRECEDE minimum standard and is appropriate for this indication.

2.2 Urgent Red Flags Requiring Expedited or Emergency Imaging

The pancreas is clinically important for emergencies, but MRI is rarely the first-line emergency investigation:

3. Preparation Reference

Universal MRI safety screening is covered in the general MRI preparation page and is not repeated here.

3.1 Anatomy-Specific Preparation Items

Fasting: 4–6 hours of fasting before pancreatic MRI/MRCP is strongly recommended. Gastric fluid and active small bowel peristalsis produce motion artefacts that severely degrade pancreatic and biliary sequences. Fasting also reduces duodenal fluid signal, which would otherwise obscure the ampullary region and periampullary structures on T2 and MRCP sequences. For MRCP specifically, fasting for 4–6 hours is the standard preparation.

Negative oral contrast (water or pineapple juice): ingestion of 200–500 mL of water 15–30 minutes before MRCP acquisition fills the duodenum and provides a natural "window" through which the ampullary region and adjacent pancreatic head are visualised on MRCP. Pineapple juice (which contains manganese from bromelain) acts as a negative oral contrast agent — its T2-suppressing effect reduces the bright duodenal signal on MRCP, improving ampullary visualisation. This is specifically useful for periampullary assessment and side-branch IPMN characterisation. Not universally used; a reasonable and low-cost technique.

Antiperistaltic agents: hyoscine butylbromide (20–40 mg IV) or glucagon (0.1–0.2 mg IV) substantially reduces duodenal peristalsis, which is the primary motion artefact source for the pancreatic head, ampulla, and distal common bile duct. Antiperistaltic agents are particularly important for MRCP quality and for T1 3D dynamic sequences covering the pancreatic head. Use is centre-dependent; contraindications include narrow-angle glaucoma and benign prostate hyperplasia for hyoscine butylbromide, and pheochromocytoma for glucagon.

Secretin preparation (conditional): if secretin-enhanced MRCP is planned, the secretin injection (0.2 μg/kg IV, maximum 16 μg) is administered during the MRCP acquisition. No specific patient preparation beyond standard fasting is required. Secretin is available as synthetic human secretin (Chirhostim, ChiRhoStim) — its availability varies significantly by country and institution. A nurse or physician must be present for administration.

Renal function: mandatory check before gadolinium injection, per standard practice.

Prior imaging: prior CT, prior MRI, prior ERCP reports, and surgical history (particularly prior pancreatic surgery, Whipple, distal pancreatectomy) must be available before scanning. Post-surgical anatomy radically changes the expected ductal morphology, enhancement pattern, and what constitutes a "normal" appearance.

3.2 Patient Positioning on the MRI System

Standard position: supine, head-first or feet-first entry depending on scanner and patient comfort. Arms elevated above the head is the preferred position for pancreatic MRI because it: (1) removes the arms from the FOV, reducing aliasing; (2) stretches the lateral abdominal muscles, reducing their interference with coil coupling; (3) allows centring at the upper abdomen without arm signal contamination. If shoulder or arm conditions preclude elevation, arms-at-sides with appropriate FOV expansion is acceptable.

Coil selection: multi-channel phased-array body matrix coil (minimum 16 channels recommended) combined with the integrated spine coil provides optimal SNR for pancreatic MRI at both 1.5T and 3T. The pancreas is a thin retroperitoneal organ (2–3 cm antero-posterior dimension in the head, 1–2 cm in the body and tail) — SNR per unit tissue is intrinsically lower than for the liver, and maximising coil channel count directly improves detection of small lesions.

Centring: isocentre at the level of the pancreatic body — approximately at the level of L1–L2, corresponding to 2–3 cm above the umbilicus. Verify on the three-plane localiser that the entire pancreas (head, neck, body, tail) and the common bile duct are within the FOV. The pancreatic tail may extend to the splenic hilum and can easily be missed if the FOV is centred too far right.

Breath-hold coaching: the single most important technologist preparation step for pancreatic MRI. The pancreas requires higher breath-hold consistency than the liver because: (1) the gland is thinner and smaller; (2) the MPD is only 2–5 mm calibre; (3) parenchymal signal changes require pixel-level reproducibility. Optimal instruction: "Breathe in, breathe out slowly, and hold at the end of expiration." Consistent expiratory position — not maximum inspiration — is the target.

Common positioning errors:

• FOV does not include the pancreatic tail (most common): the tail extends laterally to the splenic hilum; centre the FOV on the pancreatic neck/body, not on the head
• FOV centred at liver rather than pancreas: degrades resolution and coverage of the tail
• Arms at sides increasing aliasing: use phase oversampling or increase FOV in phase direction

4. Standard Protocol Design

The pancreatic MRI protocol is built around a pre-contrast core providing parenchymal assessment, duct morphology, and lesion characterisation, supplemented by dynamic post-contrast sequences for enhancement characterisation. MRCP — the unique capability of pancreatic MRI — is always included unless specifically contraindicated.

4.1 Mandatory Core Sequences

4.2 Conditional Sequences

4.3 Rationale Summary Per Sequence

T2 single-shot (HASTE/SSFSE) provides the motion-resistant baseline abdominal survey. For the pancreas, HASTE is particularly important because: (1) it is completed in a single breath-hold per slice (< 1 second per slice), making it immune to respiratory motion; (2) it depicts ductal dilation, peripancreatic fluid, and cystic lesion content clearly; (3) it allows lesion detection even in patients who cannot perform adequate breath-holds for other sequences. The limitation is lower spatial resolution compared with triggered T2 TSE, which makes it suboptimal for small ductal structures and subtle parenchymal changes.

T2 TSE with respiratory triggering provides the higher-quality T2 image for parenchymal characterisation. On triggered T2, the normal pancreatic parenchyma appears isointense to mildly hyperintense relative to the liver. T2-hyperintense foci indicate oedema, cystic transformation, or high-fluid-content lesions. Chronic pancreatitis fibrosis appears T2-hypointense. The pancreatic duct appears as a T2-bright structure that should be traceable from the tail to the ampulla. T2 TSE is the primary sequence for characterising the internal architecture of cystic lesions — septa, mural nodules, and fluid content.

MRCP 3D navigator-triggered and 2D thick-slab are the cornerstone sequences unique to pancreatic MRI. They deserve separate treatment:

3D MRCP (navigator-triggered, coronal or coronal oblique acquisition): provides high-resolution three-dimensional coverage of the entire biliary and pancreatic ductal system with isotropic or near-isotropic voxels (1.5–3 mm), enabling post-hoc multiplanar reformatting in any oblique plane. The navigator-triggered acquisition allows free breathing over several minutes, providing superior resolution compared with breath-hold MRCP. The heavily T2-weighted long TE (600–800 ms) suppresses all tissue signal, leaving only slow-moving or stationary fluid visible. The pancreatic duct, bile ducts, cystic lesions, and ductal communications are all visible.

2D thick-slab MRCP (breath-hold, single-shot, multiple angulations): provides maximum-intensity projection (MIP)-equivalent images in a single acquisition per slab, covering the full ductal system in a 3–4 second breath-hold. It is motion-resistant and provides the "overview" image analogous to an ERCP cholangiogram. Multiple angulations (straight coronal, right and left anterior oblique, right and left posterior oblique) provide complete ductal visualisation and avoid overlap with overlying fluid structures. The 2D slab provides excellent overview but cannot be reformatted — it is complementary to 3D MRCP, not a replacement.

T1 in-phase/opposed-phase performs the same function as in the liver protocol (see MRI Liver master page) with one specific addition: the normal pancreatic parenchyma has very high T1 signal on in-phase images due to its proteinaceous secretory content. Opposed-phase signal loss in the pancreas indicates fat replacement (pancreatic lipomatosis), which is associated with chronic pancreatitis, diabetes, cystic fibrosis, and advanced age. Focal signal loss on OP images in a pancreatic lesion is unusual and suggests fat-containing lesions (well-differentiated HCC metastasis; fat-containing PDAC is rare).

T1 3D fat-suppressed pre-contrast is the single most diagnostically important sequence in pancreatic MRI and is often overlooked in favour of post-contrast images. The normal pancreatic parenchyma has the highest T1 signal of any abdominal organ — brighter than the liver and spleen — due to the abundant zymogen granules in acinar cells. Any pathological process that replaces, displaces, or destroys pancreatic parenchyma (PDAC, chronic pancreatitis, autoimmune pancreatitis, pancreatitis-associated fibrosis, pancreatic necrosis) produces T1 signal loss in the affected parenchyma. This makes the pre-contrast T1 fat-suppressed sequence a highly sensitive "lesion detection" sequence: a mass that is isoattenuating on CT may be clearly T1 hypointense against the bright normal parenchyma on MRI.

DWI with multi-b-value detects diffusion restriction in hypercellular solid lesions. For the pancreas, DWI serves three functions: (1) detecting solid lesions against the T1-bright parenchymal background — PDAC and neuroendocrine tumours show restricted diffusion; (2) distinguishing solid from cystic components within complex lesions; (3) supporting PDAC characterisation in conjunction with other sequences [6, 10]. The ADC of normal pancreatic parenchyma is intermediate (approximately 1.4–2.0 × 10⁻³ mm²/s); PDAC shows reduced ADC (approximately 0.9–1.3 × 10⁻³ mm²/s), though overlap exists [10]. The PRECEDE Consortium minimum protocol includes DWI as a standard component [1].

EPI distortion is more prominent in the pancreas than in the liver because of the proximity to the air-filled duodenum and stomach. Gastric and duodenal air produces severe B0 field inhomogeneity that distorts the pancreatic head and ampullary region on EPI-DWI. Fasting and antiperistaltic agents both reduce this artefact.

T1 3D dynamic phases (arterial, portal venous, delayed) provide enhancement characterisation:

Arterial phase (approximately 20–25 seconds post-injection, or bolus-tracked): the normal pancreas enhances intensely and homogeneously in the pancreatic parenchymal phase. Hypovascular PDAC appears as a relative T1-hypointense lesion against the enhanced parenchyma — this is when PDAC is most conspicuous. Neuroendocrine tumours are hypervascular and appear T1-hyperintense in the arterial phase. The arterial phase is the most diagnostically important dynamic phase for the pancreas.

Portal venous phase (60–70 seconds post-injection): shows portal vein opacification (critical for vascular invasion assessment), peripancreatic vascular anatomy, and hepatic metastases.

Delayed phase (3–5 minutes post-injection): shows persistent or progressive enhancement in desmoplastic stroma (characteristic of PDAC, which has dense fibrous stroma that enhances progressively). Cystic lesion walls and septa show delayed enhancement. Autoimmune pancreatitis characteristically shows delayed enhancement with relative early-phase hypovascularity.

4.4 Sequence Matching and Cross-Sequence Consistency

All axial dynamic T1 3D phases must use identical geometry — same slice positions, same angulation, same FOV and matrix as the pre-contrast T1 3D acquisition. Any mismatch between phases makes lesion enhancement assessment unreliable, particularly for small lesions (< 1–1.5 cm) that may move between phases.

The 3D MRCP and the 2D slab MRCP should both be acquired before contrast injection. Post-contrast MRCP (particularly with gadoxetate in a hepatobiliary protocol that also covers the pancreas) is degraded by biliary gadolinium enhancement on hepatobiliary agents, which adds T1 signal contamination to the 3D MRCP acquisition.

For serial cyst surveillance, the MRCP 3D angulation and the axial T1/T2 prescription must be reproduced as closely as possible at each examination to ensure accurate cyst size comparison. Small differences in oblique angulation produce apparent cyst dimension changes that confound surveillance decisions. Document the MRCP angulation and the axial slice positions in the first surveillance examination.

4.5 Fat Suppression — Region-Specific Technical Considerations

Fat suppression in pancreatic MRI is applied to T1 sequences and is essential because without it, the periduodenal and peripancreatic fat has T1 signal similar to the normal parenchyma, masking the T1 contrast that is the diagnostic foundation of the examination.

Dixon fat suppression is the preferred technique for pancreatic T1 3D sequences at 3T. B0 inhomogeneity at 3T in the upper abdomen, near the air-filled duodenum and stomach, causes spectral fat saturation failure that would otherwise degrade the pre-contrast T1 quality precisely in the region of greatest diagnostic importance (the pancreatic head and ampullary region). Dixon's B0-independent mechanism provides homogeneous fat suppression throughout the pancreatic FOV.

SPAIR is acceptable at 1.5T for T1 3D pancreatic sequences. At 3T, SPAIR is the second-line option when Dixon is unavailable.

STIR is not appropriate for post-contrast T1 sequences (as discussed in the liver master page) — the inversion pulse nulls gadolinium-enhanced tissue. STIR may be used as a fat-suppressed T2 sequence in specific contexts (peripancreatic nerve invasion assessment) but is not part of the standard protocol.

T2 fat suppression: fat-saturated T2 TSE can improve pancreatic parenchymal-to-background contrast for peripancreatic inflammation and tumour infiltration assessment. Some centres use fat-saturated T2 as a primary pancreatic sequence. However, fat suppression reduces the conspicuity of peripancreatic fat infiltration (important in advanced PDAC) and is not universally applied. Both fat-suppressed and unsuppressed T2 provide useful complementary information.

4.6 Slice Positioning — Complete Technical Reference

Why Precise Slice Positioning Matters for the Pancreas

The pancreas is the most positioning-sensitive organ in abdominal MRI. Its size (15–20 cm total length, 2–3 cm antero-posterior diameter), oblique orientation (from right posterior to left anterior, angled 30–45° from the axial plane), and proximity to multiple motion artefact sources (duodenum, stomach, colon, aorta) make precise planning essential. A 5 mm error in slice positioning can place the pancreatic tail outside the FOV or misalign the pancreatic duct relative to the imaging plane, making duct diameter assessment unreliable.

Anatomical Landmarks

Pancreatic head: lies in the "C-loop" of the duodenum, to the right of the superior mesenteric vein (SMV). The uncinate process extends posterior to the SMV/SMA. The head is at approximately the level of L2.

Pancreatic neck: the narrowest part, directly anterior to the SMV-portal vein confluence. The neck is the point where the portal vein forms and is the surgical division point for Whipple procedure.

Pancreatic body: extends to the left, anterior to the aorta and the left renal vein. At the L1 level.

Pancreatic tail: extends to the splenic hilum. The tail is retroperitoneal but may have a short mesenteric attachment; it is the most mobile part of the gland and the most frequently missed in FOV planning.

Main pancreatic duct (MPD): runs the full length of the gland, visible as a T2-bright tubular structure 2–3 mm in calibre. The duct runs slightly superior in the head and joins the CBD at the ampulla of Vater.

Common bile duct (CBD): runs posterior to the pancreatic head before entering the ampulla. The intrapancreatic CBD is a critical structure that must be included in coverage.

The Planning Sequence

Pancreatic MRI requires a two-step planning approach: 1. Three-plane body localiser (standard) 2. From the coronal localiser and axial T2 single-shot, plan the dedicated pancreatic sequences along the actual orientation of the gland

Axial Slice Planning

Reference: the coronal localiser and/or sagittal T2 localiser.

Coverage extent: from the level of the hepatic veins (to include the intrahepatic bile ducts and liver dome) to below the uncinate process and the duodenal C-loop. For a comprehensive study, axial coverage from the liver dome to the lower duodenal loop is standard — approximately 20–25 cm.

Angulation: true axial (horizontal) is standard for pancreatic MRI. The pancreas lies at an angle to the true axial plane, but true axial provides the most reproducible cross-sections and the best comparison with CT. Oblique angulations perpendicular to the pancreatic duct are sometimes used for dedicated MRCP planning but not for standard axial sequences.

Phase encoding direction: A-P (anterior-posterior) for axial pancreatic sequences. This displaces motion artefacts from anterior bowel loops and the anterior abdominal wall in the A-P direction, away from the pancreas (which lies centrally). An R-L phase encoding direction would allow aortic pulsation artefacts to pass directly through the pancreatic body.

Posterior saturation band: place over the aorta and vertebral column to reduce aortic pulsation artefacts in the A-P phase direction, which otherwise produce ghosting across the pancreatic body.

Anterior saturation band (optional but useful): over the anterior stomach and bowel loops to reduce anterior motion artefacts — particularly useful when antiperistaltic agents are not used.

MRCP 3D Slice Planning

Reference: the coronal and sagittal localisers plus the axial T2 single-shot.

The key planning challenge: the MRCP must be angulated to display the MPD and CBD in the same plane simultaneously, or reformatted after acquisition to achieve this. With a 3D isotropic acquisition, any angulation can be reconstructed post-hoc. The initial acquisition is typically oriented coronally with a slight anterior tilt to align with the MPD.

Coverage: the entire biliary tree from the intrahepatic ducts to the ampulla, and the full MPD from the ampulla to the tail. Coverage must include the gallbladder. Typically 40–60 mm of coronal depth is sufficient for a slab MRCP; the 3D navigator acquisition covers a full 30–40 mm coronal slab.

Phase encoding direction for MRCP: S-I (superior-inferior) for coronal MRCP acquisitions. This places any respiratory motion ghosting in the superior-inferior direction, away from the ductal structures that course mediolaterally.

2D Thick-Slab MRCP Planning

Multiple projections are acquired from the 2D MRCP: standard coronal (straight) + 15–25° right anterior oblique + 15–25° left anterior oblique, and optionally posterior oblique projections. These overlapping angulations cover all ductal segments from multiple perspectives and prevent a single posterior or anterior structure from obscuring a ductal lesion. The projection plane should be confirmed from the axial T2 at the level of the CBD to ensure the slab passes through the full CBD length.

Oblique Axial for Pancreatic Duct Planning

When the primary clinical question is the MPD — for IPMN characterisation, chronic pancreatitis, or suspected ductal stricture — an additional thin-section (3 mm) axial oblique sequence perpendicular to the MPD course provides direct cross-section of the duct. This plane is planned from the coronal 3D MRCP by drawing the prescription perpendicular to the duct at the area of interest.

Serial Follow-Up Reproducibility

For cyst surveillance, document in the report:

• MRCP angulation in degrees from the coronal plane
• Axial coverage inferior limit (cm from umbilicus or relative to vertebral level)
• Field strength (never mix 1.5T and 3T serial measurements)

Section 4.6 — Dedicated Bibliography

Griffin N, et al. Magnetic resonance cholangiopancreatography: the ABC of MRCP. Insights Imaging. 2012;3(1):11–21. PMID: 22695995. DOI: 10.1007/s13244-011-0129-9. (Technical / Foundational) Technical reference for MRCP sequence planning and optimisation including 2D and 3D acquisition strategy, angulation methodology, and ductal coverage criteria.

Tirkes T, et al. Reporting Standards for Chronic Pancreatitis by Using CT, MRI, and MR Cholangiopancreatography. Radiology. 2019;290(1):207–215. PMID: 30398442. DOI: 10.1148/radiol.2018181353. (Moderate — Consensus reporting) Establishes slice positioning and coverage requirements for MRI/MRCP in chronic pancreatitis; defines anatomical coverage minimum standards.

Katabathina VS, et al. Magnetic Resonance Cholangiopancreatography: Current Applications and Limitations. Radiol Clin North Am. 2014;52(4):753–770. PMID: 24931183. DOI: 10.1016/j.rcl.2014.02.009. (Technical / Foundational) Comprehensive MRCP technical reference including planning strategy, angulation, contrast agent effects, and diagnostic limitations.

5. Optimisation Strategy

5.1 Artifact Reduction by Source

Respiratory motion — the dominant artefact source: the pancreas moves 1.5–3 cm during normal free breathing, and more in patients with thoracic or diaphragmatic disease. A single inconsistently held breath-hold produces a ghosted or blurred pancreas on all T1 and T2 sequences. Prevention: consistent expiratory breath-hold coaching; use antiperistaltic agents; keep each breath-hold acquisition to ≤ 18–20 seconds; use navigator-triggered 3D MRCP instead of breath-hold MRCP when breath-hold consistency is poor. If motion consistently degrades sequences, prioritise: 3D MRCP (navigator, free-breathing) → HASTE T2 (single-shot, immune to within-slice motion) → T1 dynamic (short phases with highest priority given to arterial).

Duodenal and gastric peristalsis: the duodenum wraps around the pancreatic head. Peristaltic waves produce T2-bright motion that appears on all sequences as bright arcs crossing the pancreatic head. This is the primary reason antiperistaltic agents are recommended before pancreatic MRI. Without antiperistaltic coverage, duodenal motion can simulate or obscure a periampullary or pancreatic head lesion. Mitigation: Buscopan (hyoscine butylbromide) 20 mg IV immediately before the dynamic sequences; or glucagon 0.1 mg IV.

EPI geometric distortion in DWI near the duodenum: the pancreatic head and ampullary region are adjacent to the air-filled duodenum and stomach. Air produces severe B0 field inhomogeneity that geometrically distorts the EPI DWI in this region. The pancreatic head may appear displaced by 5–10 mm from its true position on DWI, making DWI-T1 co-registration unreliable in the periampullary region. Mitigation: B0 shimming over the pancreatic FOV before DWI; use a smaller FOV centred on the pancreas; apply distortion correction (B0 field map or reverse phase encoding); fasting to reduce duodenal gas; consider multi-shot DWI (RESOLVE/MUSE) for the pancreatic head specifically.

Fat suppression failure near the duodenal loop: at 3T, the B0 field inhomogeneity from the air-tissue interface at the duodenum can produce regional fat suppression failure on the pre-contrast T1, particularly in the pancreatic head region. This mimics T1 signal change from pathology. Mitigation: Dixon fat suppression (B0-independent) is strongly preferred over SPAIR for pancreatic T1 sequences at 3T; check the fat-only image of the Dixon acquisition to confirm fat was suppressed throughout the pancreas.

Aortic pulsation artefacts: the aorta runs immediately posterior to the pancreatic body and produces T1 and T2 pulsation artefacts that propagate in the A-P phase direction and appear as ghost images of the aortic lumen overlapping the pancreatic body. Mitigation: posterior saturation band over the aorta; use cardiac gating for exquisitely sensitive sequences (rarely required); increase TR to reduce the periodicity of the ghost.

Susceptibility artefacts from biliary stents and surgical clips: metallic biliary stents (particularly stainless steel) produce severe T2* signal loss centred on the stent, obscuring the adjacent pancreatic head and CBD. Plastic biliary stents produce less artefact. At 3T, the blooming effect is 4× larger than at 1.5T. For patients with metallic biliary stents, 1.5T imaging is preferred and the MRCP findings adjacent to the stent must be reported with explicit limitation documentation.

Chemical shift artefact at the pancreatic duct wall: at narrow bandwidth, the chemical shift between the water-signal duct lumen and the surrounding fat-containing parenchyma produces a bright/dark band at the duct wall in the frequency direction. This can simulate duct wall thickening or a filling defect. Mitigation: adequate bandwidth (> 200 Hz/pixel at 3T); fat suppression on T2 TSE sequences.

5.2 Protocol Efficiency and Throughput

A complete pancreatic MRI with MRCP and post-contrast sequences typically requires 35–50 minutes at 3T, including patient preparation and breath-hold coaching. Adding secretin-MRCP extends this by 10–15 minutes.

For cyst surveillance specifically — the most common indication — an abbreviated non-contrast protocol (HASTE T2 + T2 TSE triggered + 3D MRCP + 2D slab MRCP + T1 pre-contrast 3D) can be completed in 20–25 minutes and has been shown to be non-inferior to full contrast-enhanced protocol for the surveillance decision points (worrisome features detection, cyst size measurement) [11]. This abbreviated protocol is appropriate for routine surveillance of known stable cysts.

Post-contrast sequences remain mandatory when: (1) a new solid component or mural nodule is suspected; (2) a solid lesion requires enhancement characterisation; (3) staging is the primary question; (4) autoimmune pancreatitis vs PDAC differentiation is required.

5.3 Field Strength Considerations

3T is preferred for pancreatic MRI when available. The higher intrinsic SNR enables: higher spatial resolution for small duct detection (side-branch IPMN ≤ 5 mm, small mural nodules); better DWI performance with less EPI distortion in absolute terms (though susceptibility effects are larger at 3T); faster acquisition times allowing shorter breath-holds.

Key 3T challenge for the pancreas: B0 inhomogeneity from the adjacent bowel is proportionally larger at 3T. Dixon fat suppression is essentially mandatory at 3T for all T1 sequences because SPAIR fails unpredictably near the air-filled duodenum. At 1.5T, SPAIR is more reliable.

1.5T: adequate for most pancreatic indications with optimised protocol. Better fat suppression homogeneity with spectral methods; lower susceptibility; preferred when metallic biliary stents are present. Slightly lower spatial resolution for MRCP fine detail but sufficient for most clinical questions.

The practical guidance: use 3T with Dixon for all T1 sequences; for patients with metallic biliary stents or large metallic clips, consider 1.5T for better biliary assessment.

6. Contrast Use Principles Specific to Pancreatic MRI

6.1 Non-Contrast Standard Protocol — Sufficient For

Non-contrast pancreatic MRI (T2 + T1 pre-contrast + MRCP + DWI without gadolinium) is diagnostically adequate for:

• Known stable pancreatic cyst under surveillance — multiple studies confirm non-inferiority for worrisome feature detection [11, 12]
• MRCP for duct morphology and biliary anatomy assessment
• Chronic pancreatitis structural assessment (duct morphology, parenchymal changes)
• Pancreatic divisum or ductal variant assessment
• Patients with gadolinium contraindication (renal insufficiency eGFR < 30 mL/min/1.73 m²)
• Pregnancy (non-contrast preferred; gadolinium category C)
• Exocrine function assessment with secretin-MRCP (functional information without gadolinium)

6.2 Gadolinium Indicated — Region-Specific Contexts

Gadolinium-enhanced dynamic sequences are required or strongly useful for:

• Any solid or partially solid pancreatic lesion requiring characterisation — PDAC, neuroendocrine tumour, solid pseudopapillary tumour, metastasis
• Cystic lesion with new or suspected solid component — mural nodule enhancement is a high-risk stigma per international guidelines [4, 9]
• Autoimmune pancreatitis (AIP) — delayed enhancement pattern; rim sign; mass-forming vs diffuse disease
• Pancreatic trauma — active extravasation, ductal injury, parenchymal laceration vascular assessment
• Post-treatment assessment (after chemotherapy, ablation, surgical resection)
• Suspected vascular involvement assessment for staging
• Pancreatic metastases from known extrapancreatic primary

Extracellular vs hepatobiliary agents: for the pancreas, extracellular GBCAs (gadoteridol, gadobutrol, gadoterate) are standard. Hepatobiliary agents (gadoxetate) are occasionally used when a combined liver-pancreas protocol is requested, but the hepatobiliary phase is hepatocyte-specific and adds no specific value for the pancreas itself. The arterial phase timing with gadoxetate may be suboptimal for pancreatic parenchymal assessment due to lower dose (0.025 mmol/kg vs 0.1 mmol/kg for extracellular). If the pancreas is the primary question, use an extracellular agent at full dose.

6.3 Post-Contrast Acquisition Timing

Pancreatic (arterial) phase (also called "parenchymal phase"): 20–25 seconds after start of injection, or 15–18 seconds after bolus detection trigger. The pancreas enhances maximally at this phase — hypovascular PDAC appears most conspicuous against the brightly enhanced parenchyma. This is the highest-priority phase for pancreatic imaging and must not be missed. Multi-arterial strategy (two rapid phases starting from 18 seconds) is recommended when gadoxetate is used or when the patient history includes previous missed arterial phase.

Portal venous phase: 60–70 seconds. Portal vein and hepatic parenchyma fully opacified; hepatic metastases from hypovascular primaries (PDAC metastases) are most conspicuous on portal phase.

Delayed phase: 3–5 minutes. Shows the progressive enhancement of PDAC desmoplastic stroma (which enhances slowly due to low vascularity and high fibrous content). Cystic lesion wall and septum enhancement is evaluated. Autoimmune pancreatitis shows the characteristic delayed enhancement more clearly than on earlier phases.

Note on T1 pre-contrast pancreatic parenchymal signal: the pancreatic parenchymal (arterial) phase must be compared directly with the pre-contrast T1. A lesion that appears hypointense pre-contrast and remains hypointense on the parenchymal phase is truly hypovascular. A lesion that is iso- or hyperintense pre-contrast may not be a true hypovascular PDAC even if it appears iso-enhancing on the parenchymal phase — this is why the pre-contrast T1 is mandatory and must be acquired with identical parameters before injection.

7. Reporting Essentials

7.1 Interpretation Framework

Pancreatic MRI reporting integrates findings from all sequences in a specific analytical order:

Pre-contrast T1 assessment first: evaluate the pre-contrast T1 fat-suppressed sequence as the primary lesion detection image. Normal parenchymal T1 signal should be uniformly high (brighter than the liver). Focal or diffuse T1 signal loss — even without visible mass — is a critical observation indicating parenchymal pathology. In chronic pancreatitis, diffuse T1 signal loss is the expected finding.

MRCP analysis: assess the MPD (diameter, course, regularity, filling defects), CBD (dilation, stricture, stones, communication with cysts), side branches (dilation, communication with cysts), and gallbladder. The "duct within a duct" sign (dilated MPD visible within a larger cystic lesion) indicates main-duct IPMN.

T2 and DWI for lesion characterisation: solid vs. cystic; internal architecture; restriction pattern.

Post-contrast dynamic analysis: enhancement pattern classification — hypervascular (NET, solid pseudopapillary tumour, some metastases); hypovascular (PDAC, lymphoma); peripheral/rim (abscess, pseudocyst with solid debris); progressive (desmoplastic stroma, autoimmune pancreatitis); no enhancement (cysts, intraductal mucin).

Peripancreatic assessment: portal vein and SMV involvement; SMA involvement; celiac axis; vascular invasion patterns; lymphadenopathy; liver metastases; peritoneal deposits.

7.2 Mandatory Reporting Checklist

Technical quality:

• [ ] Field strength and contrast agent documented
• [ ] Note breath-hold quality (adequate/limited/poor)
• [ ] Note antiperistaltic agent use
• [ ] Note artefacts limiting interpretation (motion, EPI distortion, stent artefact)

Pancreatic parenchyma:

• [ ] Size/atrophy: normal / focal atrophy / diffuse atrophy
• [ ] T1 pre-contrast signal: normal (bright) / reduced / focally reduced
• [ ] Parenchymal enhancement: normal / hypovascular lesion / diffuse hypoenhancement
• [ ] Fatty replacement: absent / partial / diffuse

Main pancreatic duct (MPD):

• [ ] Diameter (mm): normal (≤ 3 mm in body, ≤ 5 mm in head) / mildly dilated (3–5 mm) / significantly dilated (> 5 mm)
• [ ] Morphology: smooth / irregular / stricture / cut-off
• [ ] Filling defects (stones, mucin)

Pancreatic lesions (for each):

• [ ] Location (head/neck/body/tail/uncinate)
• [ ] Size (longest dimension; two orthogonal dimensions for cysts)
• [ ] Morphology: solid / cystic / mixed
• [ ] If cystic: septa present/absent; mural nodules present/absent; communication with MPD
• [ ] MPD calibre at lesion and upstream/downstream
• [ ] T1 signal pre-contrast
• [ ] T2 signal
• [ ] DWI: restricted / not restricted
• [ ] Enhancement pattern
• [ ] Worrisome features / high-risk stigmata (per Sendai/Fukuoka/international guidelines) [4, 9]

Biliary system:

• [ ] Intrahepatic ducts: normal / dilated
• [ ] CBD diameter (mm)
• [ ] Distal CBD: patent / stricture / filling defect
• [ ] Gallbladder: normal / stones / polyps

Peripancreatic structures:

• [ ] Portal vein and SMV: patent / thrombosis / tumour involvement
• [ ] SMA and celiac axis: clear / abutment / encasement
• [ ] Lymph nodes: normal / suspicious
• [ ] Liver: metastases present/absent
• [ ] Ascites

7.3 Structured Reporting

Reports must include: Indication (specific clinical question — cyst surveillance, solid lesion, suspected PDAC, pancreatitis, etc.); Technique (field strength, sequences, contrast agent and dose, secretin if used, quality assessment); Comparison (prior imaging date, modality, relevant findings); Findings (organised as above); Impression (direct answer to clinical question; size and characterisation of dominant finding; guideline category if applicable); Recommendations (surveillance interval per guidelines [4, 9]; biopsy; further imaging; multidisciplinary team discussion); Limitations.

For cystic lesions, explicitly state which guideline criteria are used and whether worrisome features or high-risk stigmata are present. Do not merely describe the cyst — provide the guideline-based management implication.

7.4 Incidental Findings — Clinical Decision Framework

Usually benign, document but no immediate action: simple pancreatic cysts < 1 cm with no worrisome features in a patient ≥ 60 years; pancreatic lipomatosis (fatty replacement); small liver haemangioma or cyst; gallstones without biliary complications.

Follow-up required: pancreatic cysts 1–3 cm — follow per international guidelines (Fukuoka/IAP 2024 or ACR) [4, 9]; non-dilated MPD irregularity in a context of risk factors; new diabetes with parenchymal atrophy suggesting early PDAC.

Requires urgent communication or specific action: main pancreatic duct dilation > 5 mm with no prior documentation — always warrants clinical notification; new solid or mural nodule component in a previously purely cystic lesion; dilated CBD with clinical jaundice; incidental solid pancreatic mass with arterial phase hypovascularity (strongly suspicious for PDAC — same-day clinical communication is appropriate).

8. MRI Technologist Pearls

8.1 Sequence Order Logic

The recommended sequence order for pancreatic MRI is designed to ensure the most diagnostically critical sequences are acquired when the patient is freshest and most compliant:

1. Three-plane localiser 2. HASTE/SSFSE T2 (motion-resistant; patient compliance irrelevant) 3. T2 TSE triggered (requires compliance; early while patient is alert) 4. MRCP 3D (navigator-triggered) — free-breathing; can run during T2 and T1 IP/OP planning 5. T2 TSE coronal (overview) 6. T1 IP/OP dual-echo (brief breath-hold; pre-contrast baseline) 7. DWI (pre-contrast; brief breath-hold; EPI-sensitive sequences before fatigue) 8. 2D MRCP thick-slab (multiple orientations; brief breath-holds) 9. T1 3D pre-contrast (locked reference; last pre-contrast sequence) 10. Contrast injection → Pancreatic/arterial phase → Portal venous phase → Delayed phase 11. Additional delayed if needed for AIP characterisation

The MRCP 3D navigator can run simultaneously with T1 IP/OP and DWI planning because it is navigator-triggered and does not require active patient breath-holds — this saves clock time.

8.2 Positioning Tricks

For obese patients (BMI > 35): use a smaller FOV with Dixon fat suppression; prioritise antiperistaltic agents to compensate for worse motion tolerance; consider 1.5T for better fat suppression homogeneity.

For patients who cannot hold their breath: use HASTE T2 for morphological assessment; use navigator-triggered 3D MRCP instead of breath-hold 3D MRCP; use respiratory-triggered T2 TSE; for the dynamic sequences, use compressed sensing to reduce individual phase duration to ≤ 12–14 seconds.

For DWI artefact near the duodenal loop: ask the patient to breathe out and hold 2–3 seconds after fasting and antiperistaltic agent administration; re-shim over the pancreas FOV specifically before DWI.

8.3 Fast Salvage Protocol

This 13–16 minute non-contrast protocol answers the most urgent clinical questions (cyst characterisation, duct dilation, solid lesion detection). Post-contrast dynamic phases can be added if the patient recovers compliance.

8.4 Common Avoidable Errors

9. Quality Control Checklist

• [ ] Full pancreas included (head including uncinate to tail at splenic hilum) — verify on coronal T2
• [ ] 3D MRCP: full MPD from ampulla to tail visible; CBD to intrahepatic ducts visible
• [ ] 2D MRCP: multiple angulations performed; ductal anatomy clear
• [ ] T1 pre-contrast: fat suppression homogeneous throughout pancreatic head, body, and tail
• [ ] T1 pre-contrast signal confirmed as uniformly bright in normal parenchyma (if parenchyma present)
• [ ] DWI: all b-values acquired; ADC map generated; minimal distortion over pancreatic head
• [ ] Pancreatic arterial phase: pancreatic parenchyma brightly enhanced; no portal contamination
• [ ] Portal venous phase: portal vein fully opacified
• [ ] Delayed phase: acquired at > 3 minutes post-injection
• [ ] All dynamic phases geometrically matched to pre-contrast T1 (within 2–3 mm)
• [ ] Motion artefacts assessed — each phase evaluated before leaving the dynamic acquisition
• [ ] Antiperistaltic agent administered and documented
• [ ] If secretin used: injection time documented; post-secretin acquisitions completed within 10 minutes of injection
• [ ] MRCP acquired before contrast injection (or degradation documented)
• [ ] Subtraction available if pre-contrast signal abnormally elevated

11. Evidence Gaps and Ongoing Debate

Abbreviated non-contrast MRI/MRCP for cyst surveillance: retrospective studies confirm non-inferiority of abbreviated non-contrast protocols for detecting worrisome features and guiding management decisions in known cystic lesion surveillance [11, 12]. However, formal prospective validation in large multicentre series is lacking, and no society guideline has yet endorsed an abbreviated non-contrast protocol as the standard for all cystic lesion surveillance scenarios.

MRCP vs CT for cystic lesion surveillance intervals: the ACR 2025 and Fukuoka/IAP 2024 guidelines [3, 4, 9] define surveillance intervals based on cyst characteristics and risk factors, but direct head-to-head comparative studies between CT and MRI surveillance protocols for outcome-relevant endpoints (malignant transformation detection) are lacking. MRI is preferred on theoretical grounds (soft tissue characterisation, ductal communication) and practical grounds (radiation avoidance).

Secretin-MRCP added value in cystic lesion characterisation: the evidence for secretin improving ductal communication detection in cystic lesions is equivocal — one prospective study showed modest benefit (+4.7% ductal communication detection) [13], another showed no significant difference [14]. No randomised trial has demonstrated a management outcome benefit from secretin-MRCP in routine practice.

DWI quantitative thresholds for PDAC diagnosis: published ADC threshold values for discriminating PDAC from normal parenchyma range from 0.9 to 1.3 × 10⁻³ mm²/s, reflecting platform and protocol variability rather than true biological differences. No standardised quantitative ADC criterion for PDAC is applicable across scanners without local protocol validation [10].

AI-assisted MRCP and pancreatic lesion detection: deep learning models for automated MPD measurement, cyst segmentation and growth tracking, and PDAC detection on MRI are under active development. Early-phase validation data are promising but no tool has received regulatory approval for clinical use in pancreatic MRI reporting at the time of this writing.

Secretin in the generic protocol: secretin-MRCP is recommended by some guidelines for chronic pancreatitis functional assessment and IPMN ductal communication characterisation, but availability and standardisation are highly variable. It remains a conditional rather than mandatory protocol component.

Field strength superiority: no prospective randomised study has demonstrated diagnostic superiority of 3T over 1.5T for pancreatic MRI outcomes when both protocols are optimised. Published comparative data show 3T provides higher spatial resolution for MRCP but not necessarily superior diagnostic accuracy for the primary endpoints (lesion detection, characterisation) in routine clinical practice.

12. Evidence-Based References

A. Guidelines / Consensus / Society Recommendations

B. Systematic Reviews / Meta-analyses

C. Important Prospective / Original Studies

D. Technical MRI Papers

E. Landmark Historical References