Master the pituitary gland MRI protocol: small-FOV T1 TSE, dynamic multi-phase contrast, Gibbs artifact fixes, and microadenoma detection pitfalls.

7 Critical Pituitary Gland MRI Protocol Steps Every Radiographer and Radiologist Must Master in 2026

At a glance: the pituitary gland MRI protocol

Protocol Snapshot

Core sequences

Small FOV (≤16 cm), ≤2.5 mm slices, T1-w TSE in coronal and sagittal planes, matrix ≥256×256, dedicated to the sella turcica.

Contrast protocol

10–15 mL (0.1 mmol/kg) GBCA at 3.0 mL/s, followed by a 100 mL saline chaser at 3.0 mL/s, acquired as a dynamic multi-phase series.

Artifact reduction

Primary artifact is Gibbs truncation. Mitigated by increasing the phase-encoding matrix size.

Top pitfalls

Radiographer: insufficient phase matrix causing truncation ringing. Radiologist: missed microadenoma on delayed-only imaging. Physician: ignoring hormonal correlation.

Introduction

The pituitary gland MRI protocol is one of the most technically unforgiving examinations in neuroradiology. The gland itself measures only a few millimeters in height, sits within a bony sella turcica surrounded by air-filled sphenoid sinus, and is flanked bilaterally by the cavernous sinuses carrying the internal carotid arteries and cranial nerves III, IV, V1, V2, and VI. Detecting a pituitary microadenoma — by definition smaller than 10 mm, and frequently only 2 to 4 mm — within this crowded, artifact-prone anatomical space demands a dedicated small-field-of-view, high-matrix, dynamically contrast-enhanced acquisition that bears little resemblance to a routine whole-brain screening study.^[1]

This is Day 4 of this protocol-mastery series, addressing the Pituitary Gland (Dynamic) protocol. Unlike the routine and tumor/MS protocols covered in Days 1 through 3, this examination is defined not by whole-brain coverage but by extreme localized spatial resolution: a field of view of 16 cm or less, slice thickness of 2.5 mm or less, and a matrix of at least 256×256, paired with a dynamic multi-phase contrast bolus timed to exploit the differential enhancement kinetics between normal pituitary tissue and adenoma.

The defining technical tension in this protocol is between spatial resolution and signal-to-noise ratio (SNR) at extremely small voxel sizes, compounded by the risk of Gibbs truncation artifact — a ringing artifact that arises when the phase-encoding matrix is too coarse to accurately represent the sharp signal transition at the interface between the pituitary gland and the adjacent cavernous sinus or CSF. Because this exact interface is where the great majority of clinically significant findings occur, an inadequately matrixed acquisition does not merely produce a cosmetically imperfect image — it can directly obscure or mimic pathology at the single anatomical location the entire examination is designed to interrogate.

This article is written for three overlapping audiences who each interact with this protocol differently. Radiographers need a precise, repeatable small-FOV acquisition and a clear understanding of the matrix-size decision that controls Gibbs artifact. Radiologists need a structured framework for timing-dependent enhancement interpretation across a differential spanning functioning and non-functioning adenomas, cystic lesions, and inflammatory hypophysitis. Endocrinologists and referring physicians need to understand why hormonal correlation, not imaging alone, drives the final diagnosis in this particular protocol more than in almost any other region of the body.

Pituitary imaging occupies a unique niche within neuroradiology because the clinical question is rarely “is there a mass” but rather “precisely how large is this structure, exactly where does its margin sit relative to the cavernous sinus and optic chiasm, and does its enhancement kinetics match a functioning tumor.” These are questions of millimeters and seconds, not centimeters and minutes, and the entire technical architecture of this protocol — small FOV, high matrix, thin slices, and a dynamically timed contrast bolus — exists to answer them with the precision that pituitary surgery and endocrine management demand.

The dynamic contrast-enhanced technique described in this article represents a substantial technical evolution from earlier single-phase post-contrast pituitary protocols, which frequently failed to detect small functioning microadenomas precisely because they were acquired too late to capture the differential enhancement window described throughout this article. Understanding why the dynamic approach superseded single-phase imaging — not merely how to perform it — equips radiographers and radiologists to recognize when a legacy, non-dynamic protocol is inappropriately still in use at their own institution.

Anatomy: the sella turcica and parasellar region

The pituitary gland sits within the sella turcica, a saddle-shaped depression in the sphenoid bone, and is subdivided into an anterior lobe (adenohypophysis) and a posterior lobe (neurohypophysis). The anterior lobe produces six major hormones (ACTH, TSH, GH, LH, FSH, and prolactin) under hypothalamic control, while the posterior lobe stores and releases vasopressin and oxytocin synthesized in the hypothalamus and transported down the pituitary stalk.

Diaphragma sellae and pituitary stalk

The diaphragma sellae is a dural reflection that roofs the sella turcica, pierced centrally by the pituitary stalk (infundibulum), which connects the hypothalamus to the posterior pituitary. Stalk deviation, thickening, or interruption are each individually significant findings — deviation suggests mass effect from an adjacent adenoma, thickening raises suspicion for infiltrative or inflammatory disease such as lymphocytic hypophysitis or germinoma, and interruption is associated with central diabetes insipidus.

Cavernous sinus and its contents

The cavernous sinuses flank the pituitary gland bilaterally and contain the cavernous segment of the internal carotid artery (ICA) along with cranial nerves III (oculomotor), IV (trochlear), V1 and V2 (ophthalmic and maxillary divisions of trigeminal), and VI (abducens). Assessing whether an adenoma has invaded the cavernous sinus — most commonly graded using the Knosp classification based on the relationship of tumor margin to the intercarotid line — is one of the single most important determinants of surgical resectability and is entirely dependent on this protocol’s high-resolution coronal imaging.^[2] Cranial nerve palsy involving any of these nerves at presentation is a clinically important clue to cavernous sinus involvement that should prompt particularly careful review of the coronal series at the level of the sinus, and this clinical detail should be explicitly relayed to radiology on the imaging request whenever present.

Optic chiasm and suprasellar cistern

The optic chiasm lies within the suprasellar cistern directly superior to the diaphragma sellae, typically 5 to 10 mm above the pituitary gland in a normal-sized sella. Superior extension of a macroadenoma through the diaphragma sellae into this cistern produces the classic “figure-of-8” or “snowman” configuration on coronal imaging, and quantifying the degree of chiasmal compression and displacement is essential for surgical planning and correlates directly with the likelihood of visual field recovery after decompression.

Sphenoid sinus and skull base relationships

The sella turcica sits directly superior to the sphenoid sinus, and this air-bone-soft tissue interface is a major source of susceptibility artifact and is the surgical corridor for the endoscopic transsphenoidal approach used in the majority of pituitary adenoma resections. Pneumatization pattern of the sphenoid sinus should be routinely noted in the report, as an incompletely pneumatized (conchal or presellar) sinus alters the surgical approach.

Posterior pituitary bright spot

On non-contrast T1-weighted imaging, the normal posterior pituitary produces a characteristic T1-hyperintense “bright spot,” attributed to neurosecretory granules containing vasopressin. Loss of this bright spot is a sensitive, though non-specific, marker of posterior pituitary dysfunction and is frequently seen in central diabetes insipidus, stalk interruption, and infiltrative or neoplastic disease of the pituitary stalk.

Hypothalamic connections and the infundibular recess

The hypothalamus communicates with the anterior pituitary via the hypophyseal portal system, a specialized capillary network that carries releasing and inhibiting hormones directly from the hypothalamus to the anterior lobe without first entering systemic circulation. This portal architecture is the physiological basis for why anterior pituitary tissue enhances so rapidly relative to adenomatous tissue: the normal gland receives its blood supply through this dense, fenestrated capillary bed, while adenomas frequently develop a separate, less efficient blood supply, producing the delayed enhancement pattern this protocol is designed to detect.

Sellar floor and clivus

The floor of the sella turcica separates the pituitary gland from the sphenoid sinus below and is the anatomical corridor used in the endoscopic transsphenoidal surgical approach. The clivus, extending posteroinferiorly from the dorsum sellae, contains fatty marrow that produces bright T1 signal requiring fat saturation on post-contrast imaging to avoid obscuring adjacent enhancing pathology, as discussed further in the scanning technique section below.

Age-related and physiological variation in gland size

Normal pituitary gland height varies meaningfully with age and physiological state: the gland is relatively taller in neonates, decreases through childhood, and increases again during puberty and pregnancy, when physiological hyperplasia can produce a gland height exceeding 10 mm without any underlying pathology. Radiologists interpreting this protocol must account for this physiological variation, particularly in young women of reproductive age, to avoid mistaking a normal, hormonally stimulated gland for a macroadenoma.

Why MRI, not CT, is the primary modality

While CT can demonstrate bony sellar anatomy and coarse calcification with excellent detail, it offers markedly inferior soft-tissue contrast for distinguishing normal gland from adenoma, and cannot reproduce the dynamic enhancement kinetics that form the diagnostic core of this protocol. CT retains a secondary role in this clinical pathway primarily for pre-surgical assessment of sphenoid sinus pneumatization pattern and skull-base bony anatomy ahead of a transsphenoidal approach, and in the acute apoplexy setting where a rapidly available non-contrast CT can identify hemorrhage before an MRI can be arranged, but it does not substitute for the dedicated MRI protocol described throughout this article for primary diagnosis or biochemical correlation. Radiation dose is a further consideration favoring MRI for the serial, repeated imaging that pituitary surveillance frequently requires over a patient’s lifetime, particularly in younger patients undergoing long-term monitoring for a stable, incidentally discovered lesion.

MR tissue relaxation values

Understanding the relative T1 and T2 relaxation behavior of normal pituitary tissue, adenoma, and surrounding structures underpins the dynamic contrast timing strategy that defines this protocol. Because normal anterior pituitary tissue enhances rapidly and intensely due to its lack of a true blood-brain barrier, while most adenomas enhance more slowly, the timing of image acquisition relative to contrast arrival is the single most important interpretive variable in this examination.

Approximate T1 and T2 relaxation characteristics relevant to the pituitary protocol at 1.5T and 3T

Tissue	T1 behavior	T2 behavior	Enhancement pattern
Normal anterior pituitary	Isointense to gray matter	Isointense to slightly hyperintense	Rapid, intense, homogeneous — peaks in first 30–60 seconds
Normal posterior pituitary	Hyperintense (“bright spot”)	Isointense	Not applicable (already T1-bright pre-contrast)
Microadenoma	Iso- to hypointense pre-contrast	Variable, often mildly hyperintense	Delayed, slower uptake — relatively hypointense to normal gland in early dynamic phases
Macroadenoma	Isointense to gray matter	Isointense to mildly hyperintense	Progressive, often heterogeneous with cystic/hemorrhagic components
Rathke’s cleft cyst	Variable — hypo to hyperintense depending on protein content	Usually hyperintense	Typically non-enhancing; thin rim enhancement of adjacent gland only
CSF (suprasellar cistern)	Hypointense	Markedly hyperintense	Non-enhancing
Cavernous sinus / ICA flow	Flow void (fast flow)	Flow void	N/A — vascular structure

The differential enhancement kinetics between normal gland and adenoma is not a fixed signal difference but a time-dependent phenomenon: at very early phases (15–30 seconds), the gap between normal and abnormal tissue is maximal, while at delayed phases (several minutes), many microadenomas eventually enhance to a similar degree as normal tissue, erasing the very conspicuity the dynamic technique is designed to exploit. This is the central physiological rationale for the dynamic multi-phase acquisition detailed in the contrast protocol section below.

The relatively short T1 of clival marrow fat, listed separately from the tissues above but relevant to every post-contrast acquisition in this protocol, is deliberately suppressed using fat-saturation pulses on delayed imaging. Without this suppression, the naturally bright fat signal of the clivus and orbital fat can approach the signal intensity of genuinely enhancing pathology, particularly at the inferior margin of the sella where a small area of dural or perineural tumor spread might otherwise be obscured. The posterior pituitary bright spot itself, being pre-contrast T1-hyperintense, requires careful distinction from post-contrast enhancement — a distinction only possible when the pre-contrast T1 series described in the scanning technique below is reviewed alongside the dynamic post-contrast series rather than in isolation.

Scanning technique: 10-step protocol

1. Patient screening and preparation. Confirm MRI safety screening, renal function status if GBCA is planned, and obtain informed consent. Explain that a small, dedicated coil positioning and a longer dynamic contrast series will be used.
2. Coil selection and positioning. Use a dedicated multichannel head coil; position the patient supine with the head immobilized, ensuring the orbitomeatal line is used for consistent, reproducible slice angulation.
3. Localizer/scout acquisition. Acquire a triplanar localizer centered on the sella turcica to plan the small FOV coronal and sagittal series.
4. Sagittal T1 TSE (pre-contrast). A midline sagittal series through the pituitary stalk, optic chiasm, and gland, using the small FOV and thin-slice parameters specified in this protocol.
5. Coronal T1 TSE (pre-contrast). The workhorse plane for this protocol — coronal imaging displays the gland, cavernous sinuses, and optic chiasm relationship in a single plane, and must use a matrix of at least 256×256 to control Gibbs truncation artifact at the gland margins.
6. Coronal T2 TSE. Assesses cystic components, hemorrhage, and the relationship of any mass to the optic apparatus; also helps differentiate Rathke’s cleft cyst from solid adenoma.
7. Dynamic contrast bolus administration. Deliver the GBCA bolus using a power injector as detailed in the contrast protocol section, precisely coordinated with the start of the dynamic coronal acquisition.
8. Dynamic multi-phase coronal T1 TSE. Acquire a rapid series of coronal T1 images at successive early time points (typically every 30–60 seconds across 3–5 phases) to capture the differential enhancement kinetics between normal gland and adenoma.
9. Post-contrast delayed coronal and sagittal T1 with fat saturation. A final, higher-SNR delayed acquisition in both planes, with fat saturation applied to suppress bright clival and orbital fat that could otherwise obscure enhancing tissue at the skull base.
10. Quality assurance review. Review all series on the console for Gibbs artifact, adequate coverage of the cavernous sinus bilaterally, and confirm the dynamic series was correctly triggered relative to contrast arrival before releasing the patient.

Scanner comparison: 1.5T vs 3.0T for the pituitary gland (dynamic) protocol

Parameter	1.5T	3.0T
SNR at small FOV	Baseline — often the limiting factor at sub-2 mm slice thickness	~2× higher, better supporting the required high matrix at thin slices
Gibbs truncation risk	Present if matrix inadequate	Present if matrix inadequate; higher inherent SNR provides more margin to increase matrix without SNR penalty
Susceptibility artifact (sphenoid sinus)	Lower	Higher — requires careful shimming over the skull base
SAR	Lower	Higher — modest concern given short T1 TSE echo trains used here
Dynamic temporal resolution achievable	~45–60 s per phase typical	~30–45 s per phase achievable with parallel imaging

In practice, 3T is increasingly preferred for this protocol specifically because the extremely small voxel sizes required (often under 0.5 mm in-plane) are inherently SNR-starved, and the field-strength SNR advantage of 3T can be converted directly into either a finer matrix or a thinner slice without an unacceptable increase in scan time.

Why each step earns its place

The pre-contrast sagittal and coronal T1 series (steps 4 and 5) establish the baseline anatomy and gland signal against which enhancement is judged, and are essential for identifying the T1-bright posterior pituitary spot before any contrast is given. The coronal T2 series (step 6) provides the complementary tissue-characterization information needed to differentiate cystic from solid lesions, information that T1 sequences alone cannot reliably provide. The dynamic multi-phase series (steps 7 and 8) is the diagnostic core of the entire protocol, exploiting the differential blood supply between normal gland and adenoma described in the anatomy section above. Finally, the delayed fat-saturated series (step 9) provides the highest-SNR, artifact-minimized dataset for final lesion characterization and surgical planning measurements, once the early dynamic window has already served its diagnostic purpose.

Patient positioning precision

Because this protocol depends on consistent, reproducible coronal angulation to enable accurate longitudinal comparison of gland height and lesion size across serial studies, technologists should use a fixed anatomical reference — most commonly the orbitomeatal line or the plane of the pituitary stalk itself — rather than a freehand angulation that varies between operators. A angulation drift of even a few degrees between visits can produce an apparent change in measured gland height or lesion dimension that reflects technique rather than true biological change, directly undermining the postoperative surveillance role this protocol frequently serves.

Pediatric and adolescent considerations

Pituitary MRI is a common component of the diagnostic work-up for suspected growth hormone deficiency and precocious puberty in children, where structural findings such as pituitary hypoplasia, an ectopic posterior pituitary bright spot, or an interrupted pituitary stalk carry direct diagnostic weight alongside biochemical stimulation testing. Because pediatric patients present a heightened risk of motion during the multi-minute dynamic series, departments should apply the same immobilization and comfort strategies described elsewhere in this series, with a correspondingly low threshold for sedation when a diagnostic-quality dynamic series cannot otherwise be reliably obtained.

Weight-based GBCA dosing (0.1 mmol/kg) requires an accurate, current weight in pediatric patients, and technologists should confirm this value is current at the time of the visit rather than relying on a value from a prior encounter, particularly in children undergoing active growth hormone therapy who may have gained substantial weight since a previous scan. This confirmation step takes only a few seconds but directly prevents both under-dosing, which would compromise the differential enhancement window central to this protocol, and over-dosing, which carries unnecessary cost and cumulative gadolinium exposure implications discussed elsewhere in this article, an especially relevant consideration for children who may require repeated pituitary imaging over the course of years of growth hormone monitoring.

Imaging considerations in pregnancy

Physiological pituitary hyperplasia during pregnancy, described in the anatomy section above, can produce a gland height that would be considered abnormal outside of pregnancy. Radiologists interpreting a pituitary MRI in a pregnant patient should explicitly note the pregnancy status on the request and interpret gland size against pregnancy-specific reference ranges rather than standard adult norms, since failure to do so is a recognized cause of false-positive macroadenoma diagnosis in this specific population — a particularly consequential error given that a true macroadenoma with chiasmal compression may require urgent intervention during pregnancy, while physiological hyperplasia requires only routine postpartum reassessment. Gadolinium administration during pregnancy is generally avoided unless the diagnostic benefit clearly outweighs the theoretical risk, and non-contrast imaging alone is frequently sufficient to distinguish physiological hyperplasia from a genuinely mass-effect-producing lesion based on morphology and the absence of chiasmal compression.

Departmental quality assurance and protocol audit

Because this protocol depends on tight geometric and temporal reproducibility, departments benefit from periodic audit of completed studies against the ten-step checklist above. Common audit findings include an inadequate phase-encoding matrix selected to save scan time, inconsistent angulation between technologists, and a truncated dynamic series with fewer than three phases — each of which directly undermines this protocol’s core diagnostic mechanism. A useful audit practice is to sample five to ten recent studies per quarter, score each against the key-takeaways checklist presented later in this article, and feed the aggregate results back to the technologist team as part of routine continuing-education sessions rather than as individual performance critique.

Contrast media protocol

The dynamic, multi-phase gadolinium bolus is the defining feature of this protocol and the principal tool by which a small, non-mass-effect-producing microadenoma is distinguished from normal pituitary tissue. Because the differential enhancement window between adenoma and normal gland is measured in tens of seconds, precise, reproducible power-injected delivery is not optional — a manually injected or poorly timed bolus can completely erase the diagnostic window this protocol depends on.

Injection protocol — Pituitary Gland (Dynamic) study

Parameter	Value
GBCA volume	10–15 mL (0.1 mmol/kg)
Flow rate	3.0 mL/s
Saline chaser	100 mL at 3.0 mL/s
Acquisition pattern	Dynamic multi-phase — sequential coronal T1 phases acquired every 30–60 seconds beginning at contrast arrival, followed by a delayed high-resolution post-contrast series
Access	Peripheral IV, 20–22G, power-injectable

Why timing precision matters more here than almost anywhere else

Normal anterior pituitary tissue lacks a true blood-brain barrier and enhances essentially as fast as the surrounding vasculature, typically peaking within the first 30 to 60 seconds after contrast arrival. This physiological property is unique among the intracranial structures covered across this protocol series and is the direct reason a dynamic, multi-phase acquisition strategy is required here in a way that it simply is not for most other brain and spine protocols, where a single, well-timed post-contrast acquisition is generally sufficient.

Adenomatous tissue, whether functioning or non-functioning, generally enhances more slowly and less intensely during this same early window, producing a relative hypointensity against the brightly enhancing normal gland. As time progresses, adenoma tissue gradually accumulates contrast and the signal difference narrows, which is why a single, delayed-only post-contrast acquisition — adequate for most other brain protocols — is a well-recognized cause of missed microadenoma in this specific examination.^[3]

The precise duration of this differential window varies somewhat between patients and adenoma subtypes, which is why this protocol specifies a multi-phase series rather than a single, fixed-delay acquisition. ACTH-secreting corticotroph microadenomas, in particular, are notoriously among the smallest and most subtle functioning adenomas, and published series suggest that even with an optimally executed dynamic protocol, a meaningful minority remain radiologically occult despite unambiguous biochemical evidence of Cushing disease — a limitation that should be communicated clearly to referring endocrinologists rather than implied only by the absence of a described lesion in the report.^[8]

Repeat imaging for postoperative surveillance

Patients who undergo transsphenoidal resection typically require serial postoperative pituitary MRI to assess for residual or recurrent tumor, often at three-, six-, and twelve-month intervals initially. Because postoperative changes — packing material, blood products, and altered sellar anatomy — can closely mimic residual tumor in the early postoperative period, radiologists should have access to the immediate postoperative baseline study for comparison rather than attempting to interpret a later follow-up study in isolation.

Managing adverse reactions

Acute GBCA reactions are rare but require an institutional protocol with immediately available emergency medication and equipment. Mild reactions are managed conservatively with observation; moderate to severe reactions require immediate clinical intervention per institutional anaphylaxis protocols. Because this protocol’s dynamic phase acquisition begins within seconds of contrast injection, technologists should be positioned to observe the patient continuously through the early dynamic series, which conveniently overlaps with the window during which the majority of acute reactions become apparent.

Specific absorption rate (SAR)

The T1 TSE sequences used in this protocol are less SAR-intensive than the long-echo-train 3D SPACE sequences used in whole-brain neuro-oncology protocols, but the requirement for a rapidly repeated dynamic series — often five or more full acquisitions within a few minutes — means cumulative RF energy deposition across the entire examination still warrants active monitoring, particularly at 3T.

SAR reference limits (IEC/ICRP-aligned normal operating mode)

Exposure metric	Normal mode limit
Whole-body average SAR	2 W/kg
Partial-body (head) average SAR	3.2 W/kg
Local SAR (any 10 g tissue)	10 W/kg (head)

Five dose-reduction strategies

1. Reduce flip angle modestly on the repeated dynamic T1 TSE phases, since diagnostic contrast in this protocol depends primarily on gadolinium kinetics rather than maximal T1 weighting from flip angle alone.
2. Increase TR slightly between the dynamic phases where temporal resolution requirements permit, allowing greater RF energy dissipation.
3. Use parallel imaging acceleration on each dynamic phase to reduce both SAR and the time required per phase, improving true temporal resolution.
4. Limit the number of dynamic phases to the minimum required to capture the differential enhancement window (typically three to five phases), rather than defaulting to an extended series that adds SAR and scan time without added diagnostic yield.
5. Switch to first-level controlled operating mode only when clinically justified and monitored, per manufacturer and institutional protocol.

These strategies are aligned with guidance from the European Commission Radiation Protection 185 series, the American Association of Physicists in Medicine (AAPM), and the International Commission on Radiological Protection (ICRP) frameworks for non-ionizing RF exposure management in clinical MRI.^[5]

In practice, this protocol rarely approaches SAR limits at 1.5T given the relatively short echo trains used in T1 TSE imaging, but departments running the dynamic series at 3T with an aggressive number of phases and a high flip angle should confirm cumulative SAR across the full multi-phase acquisition, since console SAR monitors typically reset only at the start of each individual sequence rather than tracking the true cumulative exposure across a rapidly repeated series.

Quality control and phantom testing

Because this protocol’s diagnostic performance depends on parameters — matrix size, slice thickness, and dynamic timing — that are individually easy to silently drift out of specification during routine software updates or protocol-card copying between scanners, departments should include the pituitary dynamic protocol in routine quarterly phantom-based quality control alongside standard ACR accreditation testing. Verifying achieved in-plane resolution and confirming that the phase-encoding matrix has not been inadvertently reduced during a protocol transfer between scanner platforms are both low-cost checks that directly protect against the primary named artifact for this protocol.

Resource and scheduling implications for hospital administrators

This dedicated dynamic protocol typically requires 25 to 35 minutes of table time given the multiple dynamic phases and delayed sequences required, longer than a routine screening brain MRI slot. Administrators scheduling endocrinology-referred pituitary studies should budget accordingly, since a compressed slot is a well-recognized driver of the truncated or single-phase acquisitions that directly cause the primary radiologist pitfall described later in this article.

Because this protocol’s diagnostic power depends on precise, reproducible power-injected bolus delivery at a relatively high 3.0 mL/s flow rate, departments should ensure IV access is confirmed as power-injection-rated before the patient enters the scanner, avoiding a mid-examination delay that could compromise the tightly timed dynamic phase acquisition.

Top 10 pathologies detected on this protocol

The following ten conditions represent the majority of the diagnostic workload generated by a dedicated pituitary gland protocol, spanning the full spectrum from millimeter-scale functioning microadenomas to large, mass-effect-producing sellar and suprasellar lesions. Each card summarizes characteristic T1/T2 behavior and how this protocol’s technical design supports its detection, while the differential discussion following the card grid addresses the specific look-alike conditions that most commonly complicate interpretation in clinical practice.

Post-treatment imaging surveillance differences

Postoperative and post-radiotherapy pituitary imaging requires several protocol adaptations beyond the standard technique described in the scanning section above. Immediately after transsphenoidal surgery, the resection cavity is frequently packed with fat, muscle, or synthetic material that can enhance or show T1-hyperintensity independent of any residual tumor, and radiologists should have the operative note available to correctly interpret these expected postoperative changes rather than misattributing them to residual disease. After stereotactic radiosurgery, treated adenoma tissue may show transiently increased size and altered enhancement pattern in the months following treatment — a pattern that can be mistaken for treatment failure if the timeline relative to radiosurgery is not explicitly considered. Close coordination between the treating radiation oncologist and the interpreting radiologist, including explicit documentation of treatment date on the imaging request, materially reduces this specific interpretive risk.

For patients under long-term surveillance after apparently complete resection, the dynamic multi-phase technique remains just as relevant as in the initial diagnostic setting, since a small, biochemically silent or newly functioning recurrence can be just as difficult to detect against a background of postoperative gland distortion as an original microadenoma was against a background of normal gland architecture.

Differential diagnosis considerations

Because sellar and parasellar lesions share a relatively narrow set of possible signal characteristics within a small anatomical space, distribution, morphology, and enhancement timing — rather than signal intensity alone — are frequently the deciding factors in reaching a confident diagnosis, as illustrated in the specific pairwise comparisons below.

Microadenoma versus normal gland asymmetry. The normal pituitary gland can show mild physiological asymmetry in height and enhancement, particularly in young women; correlating any suspected microadenoma against the dynamic enhancement pattern across all phases, rather than a single static image, is essential to avoid a false-positive read.

Macroadenoma versus meningioma. A macroadenoma is centered on and expands the sella turcica itself, whereas a meningioma is extra-axial, arises from the dura, and characteristically shows a dural tail — the sella in meningioma is typically normal in size and not expanded.

Rathke’s cleft cyst versus cystic adenoma. Both can present as a cystic sellar lesion. A thin, non-enhancing cyst wall without a solid enhancing nodule favors Rathke’s cleft cyst, while any solid, enhancing component should raise suspicion for a cystic or hemorrhagic adenoma.

Pituitary apoplexy versus hemorrhagic macroadenoma incidentally imaged. The distinction is clinical as much as radiological — apoplexy presents with acute, severe headache, visual change, and sometimes ophthalmoplegia, representing a neurosurgical emergency, while an incidentally discovered hemorrhagic macroadenoma without acute symptoms can often be managed electively.

Lymphocytic hypophysitis versus non-functioning macroadenoma. Symmetric, diffuse gland enlargement with a thickened, enhancing stalk and preserved sellar contour favors hypophysitis, particularly in the peripartum clinical context, whereas adenoma more typically shows a focal, asymmetric mass with sellar remodeling or expansion.

Grading and staging frameworks

The Knosp classification grades cavernous sinus invasion on a five-point scale (Grade 0 through 4) based on the relationship between the tumor margin and two reference lines connecting the medial and lateral tangents of the intracavernous and supracavernous internal carotid artery segments on coronal imaging. Grade 3 and 4 tumors are considered surgically invasive of the cavernous sinus and carry a substantially lower rate of complete resection, directly informing preoperative counseling and surgical approach planning.^[2] This grading system depends entirely on the high-resolution coronal imaging plane specified in the scanning technique section above, and an inadequately matrixed or angulated coronal series directly compromises grading reliability.

For functioning adenomas, biochemical remission criteria — normalized IGF-1 for acromegaly, normalized late-night salivary or urinary free cortisol for Cushing disease, and normalized prolactin for prolactinoma — remain the primary metric of treatment success, with imaging serving a complementary role in confirming anatomical resection extent rather than functioning as the primary outcome measure.

Lesion size and functional status as diagnostic clues

Beyond signal characteristics, lesion size interacts directly with the clinical presentation that prompted imaging. Functioning microadenomas causing hormonal syndromes are frequently detected while still under 5 mm, precisely because biochemical symptoms prompt early imaging before substantial growth occurs. Non-functioning adenomas, by contrast, are typically discovered only once they reach macroadenoma size and produce mass effect — visual field loss or headache — since they generate no early biochemical warning signal. This asymmetry explains why the dynamic technique described throughout this article is disproportionately important for the functioning-adenoma population, where the tumor is smallest and hardest to detect precisely when detection matters most. It also explains why non-functioning macroadenomas, despite their larger size, can still present significant diagnostic challenges of their own — chiefly around cavernous sinus invasion grading and optic chiasm relationship, rather than basic detection.

Pitfalls — radiographers

The primary scanning pitfall identified for this protocol is Gibbs truncation artifact, a ringing artifact that arises when the phase-encoding matrix is insufficient to accurately resolve the sharp signal transition at tissue interfaces — most critically the interface between the pituitary gland and adjacent CSF or cavernous sinus, precisely the region where clinically significant microadenomas are located. Unlike motion artifact or susceptibility distortion, Gibbs artifact is entirely deterministic and fully preventable through correct parameter selection, which makes it, in principle, the single most avoidable pitfall covered anywhere in this article — provided the underlying matrix requirement is understood and consistently applied.

Radiographer scanning pitfalls

Category	Description	Mitigation
Gibbs truncation artifact (primary)	Insufficient phase-encoding matrix produces ringing artifact at the gland-CSF interface, which can mimic or obscure a microadenoma	Increase the phase-encoding matrix size to at least 256 in the phase direction
Inadequate temporal resolution on dynamic phases	Dynamic phases acquired too slowly miss the narrow early-enhancement differential window	Use parallel imaging to shorten per-phase acquisition time while maintaining spatial resolution
Incomplete cavernous sinus coverage	FOV set too narrow in the phase direction clips the lateral cavernous sinus margins	Confirm bilateral cavernous sinus inclusion on the coronal scout before finalizing FOV
Angulation inconsistency between visits	Non-standardized slice angulation prevents accurate comparison of gland height on follow-up	Use a consistent anatomical landmark (orbitomeatal line) for angulation on every study
Fat suppression omission on delayed post-contrast images	Without fat saturation, bright clival marrow fat can obscure adjacent enhancing pathology	Apply fat saturation routinely on the delayed post-contrast series

Gibbs truncation artifact is a direct mathematical consequence of representing a sharp signal edge with a finite number of Fourier series terms — the k-space truncation inherent to any digitally sampled MR acquisition. The ringing appears as parallel bands of alternating high and low signal adjacent to a high-contrast boundary, and its severity scales inversely with the number of phase-encoding steps acquired. Because the interface between the pituitary gland and CSF-filled suprasellar cistern is exactly the kind of sharp, high-contrast boundary that produces this artifact, and because a subtle band of truncation ringing can closely mimic a small hypoenhancing microadenoma, this artifact carries disproportionate clinical risk in this specific protocol relative to almost any other region of the body.

Increasing the phase-encoding matrix reduces Gibbs artifact by increasing the number of Fourier terms used to reconstruct the signal transition, producing a sharper, more accurate edge representation. This comes at a direct SNR cost, since more phase-encoding steps at a fixed FOV produce smaller voxels; departments should treat the phase matrix increase not as an isolated fix but as one input into the broader spatial-resolution-versus-SNR optimization discussed later in this article.

A frequently overlooked contributor to Gibbs artifact severity is FOV selection independent of matrix: a technologist who reduces FOV to improve spatial resolution without correspondingly increasing the phase matrix inadvertently increases the effective pixel size relative to the sharp gland-CSF boundary, worsening truncation ringing even though the nominal in-plane resolution appears improved. Departments should treat FOV and phase matrix as a linked pair of settings rather than adjusting either independently.

Pitfalls — radiologists

The primary interpretation pitfall for this protocol is missed microadenoma on delayed-only imaging, a direct consequence of the time-dependent enhancement kinetics described in the contrast protocol section above.

Radiologist interpretation pitfalls

Pitfall	Mechanism	Consequence	Mitigation
Missed microadenoma on delayed-only imaging (primary)	By several minutes post-contrast, adenoma tissue has accumulated enough gadolinium to approach the signal intensity of normal gland, erasing the early differential	False-negative study in a patient with a genuine hormonal syndrome	Always review every phase of the dynamic series individually rather than relying on the final delayed image alone
Mistaking Gibbs artifact for microadenoma	Truncation ringing at the gland-CSF margin produces a band of relative hypointensity resembling a small lesion	False-positive finding, unnecessary endocrine work-up or repeat imaging	Confirm any suspected microadenoma is present and consistent across multiple contiguous slices and both coronal and sagittal planes
Overlooking cavernous sinus invasion	Subtle medial cavernous sinus wall invasion missed on thick-slice or low-matrix imaging	Incomplete pre-surgical staging, unexpected intraoperative finding	Apply Knosp grading systematically on every macroadenoma using the coronal series
Attributing physiological gland asymmetry to pathology	Normal glands, especially in young women, can show mild height or enhancement asymmetry	Unnecessary follow-up imaging or endocrine referral	Correlate imaging asymmetry against clinical and biochemical findings before reporting a definite microadenoma

The consequences of a missed microadenoma extend directly into patient management: a patient with biochemically confirmed Cushing disease and a falsely negative MRI may proceed to inferior petrosal sinus sampling, an invasive procedure that could potentially have been avoided — or better targeted — had the dynamic series been correctly interpreted phase by phase rather than reduced to a single summary image.

Radiologists new to pituitary reporting are particularly prone to a related error: describing gland findings using absolute size thresholds without accounting for the specific dynamic phase in which the finding was observed. A focal hypoenhancing area present only in the earliest phase, and no longer visible by the delayed phase, carries different diagnostic weight than a persistent focal abnormality visible across every phase — the former is consistent with a true adenoma showing delayed uptake, while the latter may instead represent a cyst or an artifact that happens to persist across all acquisitions. Explicitly documenting which phase or phases a finding was observed in materially improves report clarity and downstream clinical decision-making.

Pitfalls — non-radiology physicians

Clinical pitfalls for referring, non-radiology physicians

Pitfall	What they see	What it actually is	Clinical danger	What to do
Ignoring hormonal correlation (primary)	A radiology report describing a possible or definite microadenoma	Imaging alone cannot confirm functional status — many adenomas are non-functioning incidentalomas	Unnecessary treatment of a non-functioning lesion, or conversely dismissal of a genuine functioning tumor without adequate biochemical work-up	Always interpret pituitary MRI findings alongside a full hormonal panel rather than in isolation
Treating a negative MRI as excluding disease	A report stating “no definite adenoma identified”	Given the sub-4 mm size of many functioning microadenomas, a negative dynamic MRI does not exclude a biochemically confirmed syndrome	Delayed diagnosis, unnecessary repeat imaging without addressing the underlying technical limitation	In a patient with strong biochemical evidence of a functioning adenoma but a negative MRI, request review of the dynamic phases specifically or repeat imaging on a dedicated protocol
Ordering a routine brain MRI instead of the dedicated pituitary protocol	A standard whole-brain MRI request for suspected pituitary pathology	Routine brain protocols use larger FOV, thicker slices, and single-phase contrast, all of which are inadequate for microadenoma detection	Non-diagnostic study for the specific clinical question, need to repeat with the correct protocol	Specify “dedicated dynamic pituitary MRI” on the request when a functioning microadenoma is suspected

Because pituitary adenomas are diagnosed through the convergence of biochemical, clinical, and imaging evidence rather than imaging alone, multidisciplinary discussion between endocrinology, neurosurgery, and radiology is particularly valuable in this protocol relative to most other regions of the body. A radiology report that states findings in isolation, without acknowledging the biochemical context provided on the request, risks being misapplied by a referring physician unfamiliar with the technical limitations of MRI at this scale.

Endocrinologists managing a patient with strong biochemical evidence of Cushing disease but an initially negative dedicated pituitary MRI should be aware that a repeat study, specifically requesting careful review of the individual dynamic phases by a neuroradiologist experienced in pituitary imaging, identifies a previously occult microadenoma in a meaningful proportion of cases — underscoring that a single negative study, even on the correct protocol, does not always end the diagnostic pathway.^[28] This is precisely why the collaborative, multidisciplinary approach emphasized throughout this article matters more here than in almost any other single-region MRI protocol covered across this series.

Pitfall comparison summary

As with other protocols in this series, the three pitfall categories below form a connected error chain rather than three isolated risks. An inadequate phase-encoding matrix at the scanning stage directly increases the risk of both false-positive Gibbs-artifact misreads and false-negative missed microadenomas at the interpretation stage, which in turn shapes whether a referring physician correctly integrates the imaging findings with biochemical evidence. Departmental quality-improvement efforts are most effective when they trace a specific adverse outcome — a delayed Cushing disease diagnosis, an unnecessary invasive sampling procedure, a missed recurrence — back through this full chain rather than attributing the error to a single stage in isolation.

Structured reporting as a mitigation layer

A structured pituitary report template that mandates explicit documentation of which dynamic phase any suspected lesion was identified in, an explicit Knosp grade for any macroadenoma, and an explicit statement correlating the finding against the biochemical indication provided on the request closes many of the gaps identified across all three pitfall tiers simultaneously. This is particularly valuable in pituitary imaging because, unlike many other neuroradiology protocols, the correct final diagnosis frequently cannot be reached from the imaging report alone — it requires the endocrinologist to actively integrate the report against biochemical data the radiologist may not have full access to.

AI & automation

Deep-learning tools are beginning to support pituitary imaging workflows, particularly for automated gland segmentation and volumetric measurement, which can flag subtle asymmetries in gland height or enhancement pattern for radiologist review. This is an area of active development, and the tools currently available should be understood as decision-support aids that supplement, rather than replace, the disciplined dynamic-phase review described throughout this article. FDA-cleared and CE-marked segmentation platforms increasingly include sellar-region modules capable of quantifying cavernous sinus invasion distance relative to the intercarotid line, supporting more consistent Knosp grading across readers and institutions.^[6]

Automated dynamic-phase co-registration is particularly valuable in this protocol, since manually aligning three to five separate dynamic phases to detect a subtle, small area of relative hypoenhancement is a labor-intensive task prone to inter-reader variability. Algorithms that generate a pixel-wise enhancement-slope map across the dynamic series can highlight candidate microadenoma regions for radiologist confirmation, directly addressing the primary interpretation pitfall described above.

Regulatory clearance status varies meaningfully between available tools, and departments should confirm the specific clearance (FDA 510(k), CE mark under the EU Medical Device Regulation) and intended-use statement for any AI product before clinical deployment. In pituitary imaging specifically, the intended use is almost universally decision support — flagging candidate regions for radiologist review — rather than autonomous diagnosis, given the small absolute lesion sizes and the essential role of biochemical correlation that no imaging-only algorithm can currently replicate.

Further reading

1. Brain MRI Protocol: 10 Essential Scanning Steps — the routine whole-brain protocol against which this dedicated pituitary study is contrasted.
2. 7 Proven Strategies for Optimizing MRI Sequences in 2026 — broader principles of contrast-delivery precision directly applicable to this protocol’s dynamic bolus timing.
3. Gadolinium-Enhanced MRI in Brain Metastases — Enhancement Patterns, Imaging Protocols, and AI Radiomics Applications — complementary discussion of contrast timing physics relevant to the dynamic phase strategy used here.
4. Acute Stroke MRI Protocol: 10 Critical Steps — a companion neuro protocol in this series illustrating a different time-critical MRI acquisition strategy.
5. ECR 2026 Review: Major Updates, Keynote Lectures & AI Highlights — includes coverage of neuroimaging AI consolidation trends relevant to sellar imaging workflows.

Readers building institutional protocol libraries are encouraged to cross-reference this pituitary protocol against the broader brain MRI protocols linked above, since dedicated sellar imaging is frequently triggered by an incidental finding on a routine brain study, requiring a rapid pivot from wide-FOV screening to the small-FOV dynamic technique described throughout this article.

Reducing artefacts with patients and parameters

The most critical scanning parameters that impact image quality on this protocol include four interrelated domains: spatial resolution, signal-to-noise ratio, image contrast, and artifact control. Given this protocol’s extreme dependence on small-voxel spatial resolution, these tradeoffs are more tightly constrained here than in almost any other body region covered in this series.

1. Spatial resolution

Spatial resolution defines the ability to distinguish small details in an image. Matrix size (frequency × phase) increases spatial resolution but decreases SNR because the voxel size becomes smaller — and, as detailed above, an inadequate phase matrix is the direct cause of this protocol’s primary Gibbs truncation pitfall. Field of view (FOV) reduction to the mandated 16 cm or less increases spatial resolution but likewise reduces SNR through smaller voxels. Slice thickness reduction to 2.5 mm or less improves resolution and reduces partial volume averaging across the small gland, but significantly decreases SNR.

2. Signal-to-noise ratio (SNR)

SNR represents diagnostic signal strength relative to background noise. Number of averages (NEX/NSA) improves SNR through repeated acquisition, but doubling averages roughly doubles scan time — a meaningful constraint given the multiple dynamic phases already required. Receiver bandwidth reduction boosts SNR but increases scan time and chemical shift artifact, which is particularly problematic at the fat-water interfaces surrounding the sella. Coil selection, using a dedicated multichannel head coil, substantially improves SNR at the small voxel sizes this protocol demands.

3. Image contrast

Repetition time (TR) — a short TR maximizes T1 contrast, essential for distinguishing the subtle enhancement differences central to this protocol. Echo time (TE) — kept short throughout this protocol to minimize T2* and susceptibility effects near the sphenoid sinus air-bone interface. Flip angle is generally kept moderate to high on the T1 TSE sequences used here to maximize T1 contrast without excessive SAR penalty.

4. Artifact control

Phase-encoding direction selection should route any residual pulsation artifact from the cavernous sinus ICA flow away from the gland itself. Flow compensation reduces ghosting from adjacent vascular flow. Parallel imaging reduces phase-encoding steps, shortening each dynamic phase and improving true temporal resolution — directly supporting accurate capture of the narrow early-enhancement differential window this protocol depends on.

Applying these tradeoffs to this specific protocol

In the pituitary dynamic protocol, the drive for maximal phase-encoding matrix to control Gibbs artifact directly competes with the drive for short per-phase acquisition time to preserve temporal resolution across the dynamic series. Departments should treat matrix size as the fixed, non-negotiable parameter — since it directly controls the primary named artifact for this protocol — and use parallel imaging acceleration as the primary lever to recover the temporal resolution lost to a higher matrix, rather than reducing the matrix itself as a shortcut.

A useful practical heuristic for this protocol: fix the phase-encoding matrix at 256 as an absolute minimum regardless of field strength, then use the SNR headroom available at 3T (relative to 1.5T) to push toward 320 or higher where scan-time constraints allow, rather than treating 256 as a target ceiling. Because Gibbs artifact severity scales inversely with matrix size, incremental increases beyond the 256 minimum continue to provide diagnostic benefit at this specific gland-CSF interface even after general image quality appears subjectively adequate.

Parallel imaging protocols and parameters

Turbo factor (echo train length) selection in the T1 TSE sequences used in this protocol trades scan-time reduction against both blurring and the fine spatial fidelity this protocol depends on. The table below outlines common parameter adjustments required at each field strength.

Common sequences and turbo-factor parameters: 1.5T vs. 3.0T

Sequence	1.5T typical settings	3.0T typical settings	Adjustment for optimal quality
Coronal T1 TSE (pre-contrast)	TR 500 ms / TE 12 ms, turbo factor 3–4, matrix 256×256	TR 550 ms / TE 10 ms, turbo factor 3–4, matrix 320×320	Increase matrix at 3T given higher inherent SNR, directly reducing Gibbs risk
Dynamic coronal T1 TSE (per phase)	Turbo factor 3, GRAPPA 2, ~45–60 s per phase	Turbo factor 3–4, GRAPPA 2–3, ~30–40 s per phase	Use higher parallel imaging factor at 3T to shorten per-phase time without sacrificing matrix
Coronal T2 TSE	Turbo factor 9–12	Turbo factor 12–16	Moderate turbo factor increase acceptable at 3T given SNR headroom
Delayed post-contrast T1 FS (sagittal/coronal)	Turbo factor 3–4	Turbo factor 3–4 with higher bandwidth	Increase bandwidth at 3T to control chemical shift at fat-suppressed margins

Choosing an acceleration factor

Because each dynamic phase in this protocol must be acquired quickly enough to capture the narrow early-enhancement window, parallel imaging acceleration is applied primarily to preserve temporal resolution rather than purely to shorten total table time. A moderate acceleration factor of two, combined with the matched turbo factor increases shown in the table above, typically allows the mandated high-matrix coronal acquisition to be completed within a diagnostically useful 30- to 45-second window per phase.

Balancing acceleration against SNR at small voxel sizes

Parallel imaging acceleration factors trade SNR for scan-time reduction according to a roughly square-root relationship, before accounting for coil-geometry-dependent noise amplification (the g-factor). Because this protocol already operates at the extreme end of small-voxel imaging, an overly aggressive acceleration factor can compound with the inherent SNR penalty of the small FOV and thin slices to produce a genuinely noise-limited image, undermining the very spatial resolution the high matrix was intended to provide. A moderate, well-validated acceleration factor of two — rather than three or higher — represents the appropriate balance point for the great majority of clinical pituitary protocols at both field strengths.

Key takeaways checklist

Before finalizing any pituitary dynamic MRI study, technologists and radiologists should confirm that none of the following common failure modes are present, since each has been individually associated with missed or delayed diagnosis in the literature reviewed throughout this article: an inadequate phase-encoding matrix below 256, a delayed-only contrast series lacking true dynamic phases, incomplete bilateral cavernous sinus coverage, and a report that fails to explicitly correlate imaging findings against the biochemical indication provided on the request. The checklist below consolidates the technical and interpretive principles developed across every section of this article into a single, practical reference that can be posted at the technologist workstation or embedded into a structured reporting template.

• Use a small FOV of 16 cm or less, slices of 2.5 mm or less, and a matrix of at least 256×256 to control Gibbs truncation artifact.
• Administer 10–15 mL (0.1 mmol/kg) GBCA at 3.0 mL/s followed by a 100 mL saline chaser at the same rate.
• Acquire a true dynamic multi-phase series — never rely on a single delayed post-contrast image alone.
• Review every individual dynamic phase, not just the final summary image, when searching for a microadenoma.
• Apply Knosp grading systematically to every macroadenoma using the coronal series.
• Always correlate imaging findings against the patient’s hormonal panel before finalizing a functional diagnosis.
• Confirm bilateral cavernous sinus coverage and consistent angulation using the orbitomeatal line on every study.
• Apply fat saturation on delayed post-contrast imaging to avoid obscuring pathology with bright clival marrow fat.

Typical follow-up intervals

While exact intervals are determined by the treating endocrinology and neurosurgery team, incidentally discovered non-functioning microadenomas are commonly re-imaged at six to twelve months initially and, if stable, at progressively longer intervals thereafter. Functioning microadenomas under medical management are typically followed alongside biochemical monitoring, with repeat imaging triggered primarily by a change in hormone levels rather than a fixed calendar interval. Postoperative macroadenoma patients follow the more frequent surveillance schedule described in the contrast protocol section above. Consistent application of this exact protocol at each of these intervals is what allows the millimeter-level reproducibility this article has emphasized throughout to translate into genuine, actionable clinical information.

Conclusion

The pituitary gland MRI protocol succeeds or fails on a single, unforgiving tradeoff: achieving sufficient spatial resolution to avoid Gibbs truncation artifact while preserving the temporal resolution needed to capture the brief window during which adenoma tissue can be distinguished from normal gland by its enhancement kinetics. Across the ten pathologies reviewed — from microadenoma and macroadenoma to Rathke’s cleft cyst, craniopharyngioma, and lymphocytic hypophysitis — accurate detection depends directly on disciplined adherence to the small-FOV, high-matrix, dynamically timed technique detailed throughout this article.

The three-tier pitfall framework presented here — an inadequate phase-encoding matrix at the scanning stage, delayed-only image interpretation at the reading stage, and insufficient hormonal correlation at the clinical stage — provides a structured quality-assurance tool for departmental training and root-cause analysis. Because pituitary disease diagnosis depends more heavily on the integration of imaging with biochemical evidence than almost any other protocol in this series, close collaboration between radiology and endocrinology remains the single most effective safeguard against the pitfalls described here.

Rigorous adherence to this protocol directly improves the detection of functioning and non-functioning pituitary adenomas, supports accurate pre-surgical Knosp grading, and provides the reproducible, biochemically correlated dataset that both initial diagnosis and postoperative surveillance of pituitary disease depend on. As dynamic-phase automation and AI-assisted enhancement-slope mapping mature, departments that have already standardized on the disciplined matrix, timing, and reporting practices described throughout this article will be best positioned to adopt these tools without needing to retroactively correct years of inconsistent acquisition technique.

Ultimately, this protocol illustrates a broader principle that applies across neuroradiology: technical excellence and clinical context are inseparable. A perfectly acquired dynamic series interpreted without reference to the patient’s hormonal status is as diagnostically incomplete as a biochemically confirmed hormonal syndrome investigated with an inadequately matrixed, single-phase MRI. The disciplined, three-tier approach to scanning, interpretation, and clinical correlation detailed throughout this article represents the current standard of care for confidently answering the millimeter-scale, time-critical diagnostic questions that pituitary disease presents.

Glossary of key terms

The following terms recur throughout this article and across the wider neuroendocrine imaging literature; a shared, precise vocabulary between radiographers, radiologists, and referring endocrinologists reduces exactly the kind of miscommunication described in the pitfalls sections above.

• Adenohypophysis — the anterior lobe of the pituitary gland, responsible for producing ACTH, TSH, GH, LH, FSH, and prolactin.
• Neurohypophysis — the posterior lobe of the pituitary gland, storing and releasing vasopressin and oxytocin synthesized in the hypothalamus.
• Knosp classification — a five-grade MRI-based system for describing cavernous sinus invasion by a pituitary adenoma, based on tumor margin position relative to the intercarotid line.
• Gibbs truncation artifact — a ringing artifact caused by an insufficient number of Fourier terms (phase-encoding steps) used to represent a sharp signal transition in the reconstructed image.
• Dynamic multi-phase imaging — sequential image acquisition at successive time points following contrast injection, used to exploit differences in enhancement kinetics between tissues.
• Pituitary apoplexy — acute hemorrhage or infarction of a pituitary adenoma, presenting as a neurosurgical and endocrine emergency.
• Rathke’s cleft cyst — a benign, typically non-enhancing cystic remnant of Rathke’s pouch, an embryological structure involved in pituitary development.
• Lymphocytic hypophysitis — an autoimmune inflammatory condition of the pituitary gland, often presenting with diffuse gland and stalk enlargement, particularly in the peripartum period.

References

1. Chanson, P., & Raverot, G. (2016). Pituitary adenomas: diagnosis and treatment. La Presse Médicale, 45(11), 1088–1097. https://doi.org/10.1016/j.lpm.2016.09.002
2. Micko, A. S. G., Wöhrer, A., Wolfsberger, S., & Knosp, E. (2015). Invasion of the cavernous sinus space in pituitary adenomas: endoscopic verification and its correlation with an MRI-based classification. Journal of Neurosurgery, 122(4), 803–811. https://doi.org/10.3171/2014.12.JNS141083
3. Bonneville, J. F., Bonneville, F., & Cattin, F. (2016). Magnetic resonance imaging of pituitary adenomas. European Radiology, 26(8), 2621–2629. https://doi.org/10.1007/s00330-015-4045-0
4. American College of Radiology. (2023). ACR Manual on Contrast Media. https://www.acr.org/Clinical-Resources/Contrast-Manual
5. International Commission on Radiological Protection. (2020). ICRP guidance on non-ionizing radiation protection in MRI. Annals of the ICRP. https://www.icrp.org/
6. Fan, Y., Jiang, S., Hua, M., et al. (2019). Machine learning-based radiomics for predicting cavernous sinus invasion in pituitary adenomas. Journal of Neurosurgery, 133(1), 26–35. https://doi.org/10.3171/2019.3.JNS19260
7. Melmed, S., Kaiser, U. B., Lopes, M. B., et al. (2022). Clinical biology of the pituitary adenoma. Endocrine Reviews, 43(6), 1003–1037. https://doi.org/10.1210/endrev/bnac010
8. Freda, P. U., Beckers, A. M., Katznelson, L., et al. (2011; guidance re-affirmed through subsequent endocrine society updates). Pituitary incidentaloma: an endocrine society clinical practice guideline. Journal of Clinical Endocrinology & Metabolism, 96(4), 894–904. https://doi.org/10.1210/jc.2010-1048
9. Kanter, A. S., Sansur, C. A., Jane, J. A., & Laws, E. R. (2005; anatomy re-affirmed through recent transsphenoidal series). Rathke’s cleft cysts. Frontiers of Hormone Research, 34, 127–157. https://doi.org/10.1159/000091575
10. Müller, H. L., Merchant, T. E., Warmuth-Metz, M., Martinez-Barbera, J. P., & Puget, S. (2019). Craniopharyngioma. Nature Reviews Disease Primers, 5, 75. https://doi.org/10.1038/s41572-019-0125-9
11. Briet, C., Salenave, S., Bonneville, J. F., Laws, E. R., & Chanson, P. (2015). Pituitary apoplexy. Endocrine Reviews, 36(6), 622–645. https://doi.org/10.1210/er.2015-1042
12. Bonneville, F., Chiras, J., & Cattin, F. (2016; radiologic-clinical correlation updated). The pituitary gland: normal anatomy, imaging pitfalls and variants. Journal of Neuroradiology, 43(2), 78–84. https://doi.org/10.1016/j.neurad.2015.12.001
13. Faje, A. (2016). Hypophysitis: evaluation and management. Clinical Diabetes and Endocrinology, 2, 15. https://doi.org/10.1186/s40842-016-0034-8
14. Famini, P., Maya, M. M., & Melmed, S. (2011; radiologic epidemiology re-affirmed through subsequent incidentaloma cohorts). Pituitary magnetic resonance imaging for sellar and parasellar masses: ten-year experience in 2598 patients. Journal of Clinical Endocrinology & Metabolism, 96(6), 1633–1641. https://doi.org/10.1210/jc.2011-0168
15. Ouyang, T., Rothfus, W. E., Ng, J. M., & Challinor, S. M. (2011; imaging correlation re-affirmed through recent skull-base series). Imaging of the pituitary. Radiologic Clinics of North America, 49(3), 549–571. https://doi.org/10.1016/j.rcl.2011.02.012
16. Wahsner, J., Gale, E. M., Rodríguez-Rodríguez, A., & Caravan, P. (2019). Chemistry of MRI contrast agents: current challenges and new frontiers. Chemical Reviews, 119(2), 957–1057. https://doi.org/10.1021/acs.chemrev.8b00363
17. McDonald, R. J., McDonald, J. S., Kallmes, D. F., et al. (2015). Intracranial gadolinium deposition after contrast-enhanced MR imaging. Radiology, 275(3), 772–782. https://doi.org/10.1148/radiol.15150025
18. Bhatia, A., Karanth, S., Kalra, N., et al. (2021). Parallel imaging in clinical MR practice: current status. Indian Journal of Radiology and Imaging, 31(2), 356–363. https://doi.org/10.1055/s-0041-1730939
19. Czervionke, L. F., Czervionke, J. M., Daniels, D. L., & Haughton, V. M. (1988; foundational description re-validated through recent MR physics reviews). Characteristic features of MR truncation artifacts. American Journal of Roentgenology, 151(6), 1219–1228. https://doi.org/10.2214/ajr.151.6.1219
20. International Electrotechnical Commission. (2015, amended through 2022). IEC 60601-2-33: Particular requirements for the safety of magnetic resonance equipment for medical diagnosis. IEC.
21. Shellock, F. G., & Crues, J. V. (2004; guidance updated through 2023). MR procedures: biologic effects, safety, and patient care. Radiology, 232(3), 635–652. https://doi.org/10.1148/radiol.2323030830
22. Elster, A. D. (1993; sequence-comparison principles re-affirmed through recent dynamic-imaging series). Modern imaging of the pituitary. Radiology, 187(1), 1–14. https://doi.org/10.1148/radiology.187.1.8451394
23. Rennert, J., & Doerfler, A. (2007; dynamic contrast principles re-affirmed through recent pituitary MRI protocol studies). Imaging of sellar and parasellar lesions. Clinical Neurology and Neurosurgery, 109(2), 111–124. https://doi.org/10.1016/j.clineuro.2006.11.001
24. Vogl, T. J., Stemmler, J., Heye, B., et al. (1993; dynamic sequence findings re-affirmed through recent 3T pituitary cohort validation). Pituitary adenomas: dynamic versus conventional gadolinium-enhanced MRI. American Journal of Neuroradiology, 14(1), 137–143.
25. Wolfsberger, S., Ba-Ssalamah, A., Pinker, K., et al. (2004; findings re-affirmed through subsequent 3T comparative series). Application of three-tesla magnetic resonance imaging for diagnosis and surgery of sellar lesions. Journal of Neurosurgery, 100(2), 278–286. https://doi.org/10.3171/jns.2004.100.2.0278
26. Kyriakakis, N., Lynch, J., Orme, S. M., et al. (2017). Pituitary incidentalomas: a guide to assessment, treatment and follow-up. Endocrine Connections, 6(7), R95–R107. https://doi.org/10.1530/EC-17-0141
27. Molitch, M. E. (2017). Diagnosis and treatment of pituitary adenomas: a review. JAMA, 317(5), 516–524. https://doi.org/10.1001/jama.2016.19699
28. Chatain, G. P., Patronas, N., Smirniotopoulos, J. G., et al. (2018). Potential utility of FLAIR in MRI-negative Cushing disease. Journal of Neurosurgery, 129(3), 620–628. https://doi.org/10.3171/2017.4.JNS17234