MRI Brain – Generic Standard Protocol

1. Executive Summary

Brain MRI is the dominant imaging modality for the evaluation of intracranial pathology across the entire spectrum of neurological disease. Its superiority over CT rests on multiparametric tissue characterisation, absence of ionising radiation, absence of posterior fossa beam-hardening artefact, and the ability to detect pathology — particularly early ischaemia, microhaemorrhage and demyelination — that may be invisible on CT [1].

The standard brain MRI protocol is intentionally designed as a broad-sensitivity survey: it is optimised to detect a wide range of pathology without prior lesion localisation. This is its core strength and its fundamental limitation — it is a diagnostic compromise between depth and breadth.

1.1 Core strengths

• Early acute ischaemia detection using DWI, often within minutes of onset.
• Microhaemorrhage and cortical siderosis detection using SWI.
• White matter disease characterisation using FLAIR.
• Improved posterior fossa and brainstem assessment compared with CT because there is no bone-hardening artefact.
• Repeatable, non-ionising imaging suitable for serial monitoring.
• Multiparametric evaluation: each sequence interrogates different tissue properties.

1.2 Intrinsic limitations of the generic protocol

The generic protocol is not optimised for small lesions requiring dedicated thin-slice acquisition, such as pituitary microadenoma below 3 mm, cochlear nerve or labyrinthine pathology. It does not provide haemodynamic, metabolic or functional information and remains highly dependent on patient cooperation. Protocol completeness is also a direct trade-off against scan time and throughput.

Posterior fossa quality may be inferior to supratentorial assessment in routine 2D acquisitions, and non-contrast imaging is insufficient for disease categories requiring enhancement characterisation. Any specific clinical suspicion — epilepsy, multiple sclerosis, pituitary disease, vestibular schwannoma, intracranial mass, dementia or cerebrovascular malformation — warrants protocol modification in a dedicated child protocol.

2. Main Clinical Indications

The ACR Appropriateness Criteria [1] and ESR iGuide [2] provide the evidence framework for referral justification. The standard protocol is a reasonable first-line investigation when the clinical question is broad, non-localising or not yet specific enough to justify a dedicated protocol.

2.1 Standard Indications

Headache. MRI is indicated for new, progressive or atypical headache when red flags are present or CT is non-contributory. Yield for significant structural pathology in chronic featureless headache is low, commonly estimated below 1–3%, but non-negligible; incidental vascular findings, white matter hyperintensities and small aneurysms may be encountered in a clinically relevant proportion [3]. The non-contrast standard protocol is sufficient for most headache indications; detailed decision points belong to the headache child page.

Dizziness and vertigo. Persistent unexplained vestibular symptoms, atypical patterns or central features mandate brain MRI to exclude posterior fossa mass, cerebellar infarction or demyelination. The standard protocol plus dedicated posterior fossa attention is the starting point; an IAC protocol is added when vestibular schwannoma or inner ear pathology is suspected.

First seizure and epilepsy screening. Brain MRI is the recommended first-line structural investigation after unprovoked seizure. The standard protocol is the minimum acceptable baseline; drug-resistant epilepsy requires a dedicated high-resolution epilepsy protocol with hippocampal-oriented imaging.

Cognitive impairment and memory decline. The standard protocol screens for potentially reversible causes such as hydrocephalus, subdural collections and vascular disease, and characterises atrophy patterns. Dedicated dementia workup augments the protocol with hippocampal volumetry, ASL perfusion or other advanced methods where indicated.

Non-focal neurological symptoms. Fatigue, diffuse paraesthesiae, mild gait difficulty or non-specific slowing without lateralising signs represent a common referral category. In this setting, the standard protocol functions as an appropriate first examination and exclusion tool.

Chronic condition monitoring. Established MS, post-stroke follow-up and known small vessel disease may use the standard protocol as a baseline, with disease-specific modifications defined in child pages.

Incidental finding follow-up. Prior incidental findings such as cysts, small aneurysms or white matter hyperintensity burden often require interval monitoring. The standard protocol is generally sufficient unless the specific finding demands dedicated sequences.

2.2 Urgent Red Flags Requiring Expedited or Emergency Imaging

Alert: The following presentations are not appropriate for routine outpatient scheduling and usually require urgent or emergency imaging assessment.

3. Preparation Reference

3.1 Brain MRI-Specific Preparation Items

Brain MRI is particularly vulnerable to small metallic or susceptibility-generating items located around the face, ears, oral cavity and skull base. These items may have little relevance for other anatomical regions but can seriously degrade neuroradiological assessment.

• Hearing aids: must be removed before entering the controlled MRI area. If left in place, they may stop working and may generate susceptibility artefact over the temporal bones, petrous pyramids and posterior fossa; in IAC protocols this can invalidate the examination.
• Removable dental prostheses: removable plates, bridges and dentures should be removed. Fixed crowns and implants cannot be removed, but their presence should be anticipated because they may degrade inferior frontal, anterior temporal and skull base sequences, especially DWI and SWI. Sometimes the artifact can propagate widely throughout the entire study volume.
• Facial piercings: nose studs, eyebrow bars, tongue piercings and lip rings should be removed when physically possible. Non-removable piercings should be documented because they may affect SWI, DWI and orbital/anterior brain evaluation.
• Eye makeup and metallic cosmetics: mascara, eyeliner and eyeshadow may contain metallic pigments such as iron oxides. Patients should be instructed at scheduling to arrive without makeup when orbits, skull base or anterior temporal structures are relevant. Makeup should be removed with removal wipes, which should be available in the department. Metallic makeup and eyeliner tattoos can cause eyelid skin burns.
• Hair accessories: hairpins, metallic clips, extensions or wigs with metallic components must be removed. They are common missed causes of susceptibility artefact and may also represent a safety risk.
• Glasses: remove before entering the scanner environment.
• VP shunt history: brain MRI requests should explicitly identify programmable shunts, because post-MRI valve verification or reprogramming must be arranged according to local workflow.
• Prior neurosurgery: operative history, shunt hardware, embolisation material, clips and stimulator leads should be known before scanning. Detailed compatibility evaluation belongs to the general MRI safety page, but the brain protocol must remain alert to the local imaging consequences of these devices.

3.2 Patient Positioning on the MRI System

The standard position is supine, head-first, with the head centred symmetrically inside a dedicated multichannel head coil or head/neck coil. The patient should be aligned so that the nose points straight upward, both ears are equidistant from the coil sides and there is no head rotation. Even mild lateral rotation can degrade shimming symmetry, reduce reproducibility in follow-up studies and increase motion from discomfort.

Use foam pads bilaterally between the head and coil to reduce subtle rocking motion. A small support under the neck may improve comfort, but excessive flexion or extension should be avoided. Knee support may reduce lumbar discomfort and indirectly reduce head motion during longer acquisitions.

The coil should not compress the face or nose, and the patient must have the alarm bell in hand before the table moves into the bore.

For anxious patients, confirm comfort before starting long 3D sequences such as SWI, 3D FLAIR or MPRAGE. Do not begin these sequences immediately after repositioning or reassurance; allow the patient to settle first. For best image quality, the technologist should verify the first localiser and early sequence images for head tilt, incomplete vertex/foramen magnum coverage and gross motion before committing to the full protocol. For sequence-level protocol optimisation, vendor terminology and artefact management, see the dedicated MRIninja page MPRAGE / 3D T1 Magnetisation-Prepared GRE Sequence.

4. Standard Protocol Design

The standard brain MRI protocol is built around five mandatory sequences forming the diagnostic core, with conditional and optional sequences added by clinical indication.

4.1 Mandatory Core Sequences

4.2 Conditional Sequences

4.3 Rationale Summary Per Sequence

DWI + ADC

Detects restricted water diffusion. It is clinically irreplaceable for acute ischaemia, pyogenic abscess, epidermoid and hypercellular neoplasms. DWI and ADC must always be interpreted together to avoid T2 shine-through errors. When performed as the first sequence, it provides rapid image acquisition that can be reviewed immediately, allowing the operator to identify unexpected abnormalities and, if necessary, adapt the remaining study protocol accordingly. Tech Tips -->This sequence is considerably louder than most other brain MRI acquisitions. When used first, it may surprise or alarm the patient. Briefly warn the patient before starting, explaining that the first minute will be particularly noisy.

Sagittal T1-weighted

Provides midline overview, corpus callosum, posterior fossa and craniocervical junction anatomy. It detects lipid-containing lesions, subacute haemorrhage with methemoglobin T1-bright signal and macroscopic structural anomalies. It is fast and provides an orientation reference for subsequent planning. Anatomical definition in the sagittal plane may also be adequately provided by other sequences acquired in this orientation, depending on the protocol design.

Axial (or Coronal) T2-weighted

Primary parenchymal sensitivity sequence. T2 hyperintensity reflects oedema, gliosis, demyelination and many neoplastic processes. In Axial orietation is particularly important for brainstem and posterior fossa assessment because it is less affected by FLAIR-type CSF pulsation artefacts.

FLAIR

FLAIR is T2-weighted imaging with CSF nulling. It renders periventricular, juxtacortical and subarachnoid lesions conspicuous by suppressing competing CSF signal. It is complementary to T2 and does not replace it. With current technology, many high-quality centres perform FLAIR, when patient cooperation allows, as a 3D acquisition in the sagittal plane, obtaining isotropic voxels often around 1 mm in each dimension and enabling high-quality multiplanar reconstructions.

SWI

A 3D gradient-echo sequence exploiting magnitude and filtered phase information to maximise susceptibility sensitivity. It detects microhaemorrhages, cortical siderosis, cavernomas, developmental venous anomalies, calcification-versus-haemorrhage differentiation and cerebral iron deposition. It is more sensitive than T2* GRE for these tasks [5,6].

4.4 Pre- and post-contrast T1 sequence matching

A fundamental principle of contrast-enhanced brain MRI is that the post-contrast T1 sequence should be acquired using parameters identical to those of the pre-contrast T1 acquisition — same sequence type, same slice geometry, same angulation, same slice thickness, same matrix, and same FOV. This matching is essential for two distinct clinical reasons. First, reliable enhancement detection requires direct voxel-to-voxel comparison between pre- and post-contrast images: a signal increase that appears convincing on the post-contrast image alone may represent true gadolinium enhancement, but it may equally represent a T1-shortening effect from subacute haemorrhage, proteinaceous content, melanin, calcification, or fat — all of which produce spontaneous T1 hyperintensity independent of contrast administration. Without an identical pre-contrast reference, these findings are indistinguishable from true enhancement and may lead to significant diagnostic error. Second, in the presence of known or suspected spontaneous T1 hyperintensities — subacute haematomas, haemorrhagic metastases, melanotic lesions, lipomatous components, or post-treatment changes — the pre-contrast T1 is not merely useful but diagnostically indispensable, as enhancement assessment without it is unreliable by definition. When pre- and post-contrast sequences are geometrically identical, digital subtraction of the two datasets is also feasible, further increasing the conspicuity of subtle enhancement patterns such as thin leptomeningeal enhancement or small dural metastases. This is contingent on the patient remaining completely still between the two acquisitions: any head movement between the pre- and post-contrast T1 — even of a few millimetres — introduces spatial misregistration that renders digital subtraction unreliable, producing edge artefacts and false signal differences that may simulate or obscure enhancement. When patient cooperation is adequate and geometric matching is confirmed, subtraction images are particularly valuable for the detection of subtle leptomeningeal and pachymeningeal enhancement, thin peripheral rim enhancement of small metastases, early post-operative enhancement patterns, and any enhancing lesion adjacent to spontaneously T1-hyperintense structures where visual comparison alone is insufficient to confirm true gadolinium uptake. For all of the above reasons, it is strongly recommended that an intravenous access be established before the examination begins in all cases where contrast administration is anticipated or cannot be excluded, and that injection be performed using an automated power injector rather than manual bolus. Automated injection ensures a reproducible injection rate, a consistent contrast bolus profile, and — critically — allows the patient to remain completely still inside the bore without any physical interaction from the operator during the acquisition: the injector is programmed before the sequence starts and delivers the contrast agent and saline flush autonomously, making the entire injection process transparent to the patient and minimising the risk of motion between pre- and post-contrast acquisitions. Manual injection, by contrast, requires the operator to enter or interact with the patient, inevitably introducing the risk of head movement at the moment of injection and compromising the geometric matching essential for reliable subtraction imaging

In high-throughput clinical settings where protocol time is a limiting factor, strict pre/post matching may be relaxed provided that the clinical indication does not involve haemorrhagic lesions, spontaneous T1 hyperintensities, or situations requiring subtraction imaging. In these contexts, a rapid 2D axial T1 post-contrast may be acquired without a dedicated pre-contrast counterpart, accepting the diagnostic limitation that enhancement cannot be formally confirmed in the absence of a baseline — a trade-off that must be consciously made and, where relevant, acknowledged in the radiology report.

4.5 Fat suppression in brain MRI — indications, advantages, and limitations

Fat suppression is not a routine component of standard brain MRI sequences and is generally not applied to the core protocol acquisitions — T1 TSE, T2 TSE, FLAIR, SWI — in their standard non-contrast configuration. However, understanding when fat suppression adds diagnostic value and when it introduces unacceptable trade-offs is essential for protocol optimisation in specific clinical contexts.

The rationale for fat suppression in neuroimaging derives from two distinct needs: the elimination of competing high T1 signal from fat-containing structures that may obscure or simulate pathological enhancement on post-contrast T1 sequences, and the suppression of chemical shift artefact at fat-water interfaces — most relevant at the skull base, the orbits, the scalp-calvarium interface, and in the vicinity of fat-containing lesions such as lipomas, dermoid cysts, and teratomas.

The main fat suppression techniques available in clinical brain MRI and their specific characteristics are the following:

• Spectral fat saturation (ChemSat / SPIR / SPAIR): applies a frequency-selective RF pulse tuned to the resonance frequency of fat protons (approximately 3.5 ppm from water, corresponding to 220 Hz at 1.5T and 440 Hz at 3T) to saturate fat signal before the imaging sequence. It is the most widely used technique for post-contrast T1 fat-suppressed sequences in the orbits, skull base, and spine. Its main advantage is that it does not significantly prolong acquisition time and preserves T1 contrast of enhancing lesions. Its principal limitation is sensitivity to B0 field inhomogeneity: in regions where the static field is not perfectly homogeneous — which includes the skull base, the anterior temporal fossae, and areas adjacent to metallic implants or dental hardware — the fat saturation pulse may be mistuned, resulting in incomplete fat suppression in some areas and inadvertent water suppression in others, producing characteristic regional signal dropout that can simulate or obscure pathology. This limitation is significantly more pronounced at 3T, where B0 inhomogeneity is inherently greater than at 1.5T.
• STIR (Short Tau Inversion Recovery): uses an inversion recovery preparation pulse with a short TI timed to null the longitudinal magnetisation of fat at its zero-crossing point (TI approximately 150–180 ms at 1.5T; approximately 180–210 ms at 3T). STIR provides robust fat suppression that is largely independent of B0 homogeneity and is therefore more reliable than spectral fat saturation in regions of field inhomogeneity. However, STIR cannot be used in combination with gadolinium-based contrast agents for enhancement assessment: the inversion recovery preparation nulls signal based on T1 value, and gadolinium shortens the T1 of enhancing tissues to values that may overlap with the fat null point, causing enhancing lesions to appear suppressed rather than bright — a critical diagnostic pitfall. STIR is therefore reserved for non-contrast acquisitions.
• Dixon technique (two-point, three-point, or multi-point): acquires images at two or more echo times to exploit the phase difference between water and fat protons, mathematically separating water-only and fat-only images in post-processing. Dixon provides the most homogeneous fat suppression of all available techniques, is robust to B0 inhomogeneity, and can be combined with both non-contrast and post-contrast sequences without the limitations of STIR. It is increasingly available on modern scanners and is the preferred fat suppression method for post-contrast T1 sequences in challenging anatomical regions such as the skull base and orbits. Its main disadvantages are longer acquisition time compared to spectral fat saturation and slightly increased complexity of post-processing reconstruction.

In routine standard brain MRI, fat suppression on post-contrast T1 sequences is indicated when the clinical question involves the orbits, the optic nerves, the cavernous sinuses, the skull base, the meninges adjacent to calvarial fat, or any suspected fat-containing lesion. In these contexts, a fat-suppressed post-contrast T1 — preferably using Dixon or SPAIR at 3T — substantially increases the conspicuity of enhancing pathology by eliminating the competing high T1 signal from adjacent orbital fat, bone marrow, and subcutaneous fat. Conversely, routine application of fat suppression to all post-contrast brain T1 sequences in the absence of a specific indication is not recommended, as it prolongs acquisition time, introduces the risk of inhomogeneous suppression artefacts, and provides no diagnostic benefit when the clinical question involves purely intraparenchymal pathology remote from fat-containing structures.

On non-contrast sequences, fat suppression has a limited role in standard brain MRI. The one context where it provides unambiguous value is in the characterisation of fat-containing lesions: a T1 hyperintense lesion that loses signal on a fat-suppressed T1 sequence confirms lipid content and narrows the differential diagnosis to lipoma, dermoid cyst, or teratoma. Outside this specific indication, fat suppression on non-contrast brain sequences introduces unnecessary complexity without diagnostic gain in the majority of clinical referrals.

5. Optimisation Strategy

5.1 Artifact Reduction by Source

Motion artefact is the primary cause of non-diagnostic brain MRI. Invest time in patient counselling and positioning; rushed setup leads to compensatory motion. Short sequences such as sagittal T1 and DWI can establish patient cooperation before longer sequences such as FLAIR, SWI and MPRAGE. Parallel imaging with acceleration reduces sequence duration. For 3D sequences, motion corrupts the entire volume, so never start SWI, MPRAGE or 3D FLAIR until the patient is settled.

Dental and metallic susceptibility artefact is most severe in EPI-DWI and SWI. Metal in the oral cavity propagates distortion particularly at the anterior temporal lobes, inferior frontal lobes and skull base. If susceptibility obscures a critical region, RESOLVE/readout-segmented DWI can reduce distortion, while T2 TSE remains more reliable than EPI-based sequences for skull base and inferior temporal assessment. For sequence-level protocol optimisation, vendor terminology and artefact management, see the dedicated MRIninja page Spin Echo DWI / Non-EPI DWI Sequence.

Posterior fossa FLAIR artefact is a major interpretative pitfall. Pulsatile CSF flow in the fourth ventricle and aqueduct may create high-signal artefact simulating leptomeningeal or intraventricular pathology. Always correlate with T2 before attributing posterior fossa FLAIR signal to disease.

Pulsatile CSF flow artefact Pulsatile CSF flow artefact results from the rhythmic motion of cerebrospinal fluid driven by the cardiac cycle: with each systole, arterial pulsation displaces CSF in a bidirectional pattern predominantly through the cerebral aqueduct, the fourth ventricle, and the basal cisterns, introducing phase shifts that are misregistered along the phase encoding direction as ghost copies of CSF-containing structures. The artefact affects all standard brain MRI sequences but is most clinically significant on FLAIR, where the long TR and TE maximise phase error accumulation and incomplete CSF suppression in the posterior fossa produces focal areas of high signal that closely mimic leptomeningeal pathology or subarachnoid haemorrhage — a pitfall documented in up to 73–86% of routine brain FLAIR acquisitions and representing the single most important posterior fossa FLAIR interpretation trap; systematic correlation with T2 TSE is mandatory before attributing any posterior fossa FLAIR hyperintensity to pathology. On T2 TSE, pulsatile flow in the aqueduct produces signal heterogeneity that should not be misinterpreted as aqueductal stenosis. On SWI, phase errors from pulsatile venous flow may simulate focal susceptibility foci or venous thrombosis adjacent to dural sinuses. On DWI-EPI, cardiac-synchronous brainstem motion can produce apparent diffusion restriction mimicking acute ischaemia in the posterior fossa. The artefact is intrinsically more prominent at 3T than at 1.5T and more severe with 2D than 3D acquisitions; partial mitigation is achievable with spatial saturation bands, phase encoding direction swap, and 3D FLAIR substitution, while cardiac gating remains the definitive solution but is rarely applied in routine clinical brain MRI due to acquisition time overhead.

Chemical shift artefact is relevant at the skull base and fat-containing lesions. Wider receive bandwidth reduces chemical shift displacement. This matters more at 3T, where fat-water frequency separation is approximately double that at 1.5T.

Vascular pulsation artefacts Arterial pulsation artefacts arise from the periodic motion of blood within intracranial vessels synchronised with the cardiac cycle, and manifest through two distinct mechanisms that must be recognised separately. The first is the flow void, a signal loss phenomenon occurring in vessels with rapid, coherent flow — most prominently the internal carotid arteries, the basilar artery, the middle cerebral artery trunks, and the major dural sinuses — where protons moving through the slice during the TE interval dephase or exit the excited volume before signal readout, producing characteristic signal-free tubular or rounded structures on T2 and FLAIR images. Flow voids are a normal and expected finding in patent high-flow vessels and their absence or asymmetry may in fact indicate pathological flow reduction, thrombosis, or severe stenosis; their presence should therefore be actively noted as a positive sign of vessel patency rather than dismissed as artefact. The second mechanism is the pulsatile ghost artefact, which occurs when periodic motion of vessel walls and surrounding blood column introduces phase shifts that are inconsistently encoded across the multiple TR repetitions of a 2D TSE acquisition; because the motion is cardiac-synchronous but not synchronised to the MRI pulse sequence timing, the resulting phase errors are distributed across k-space and reconstructed as multiple equidistant ghost copies of the pulsating structure, displaced at regular intervals along the phase encoding direction with progressively decreasing intensity. In brain TSE sequences, this manifests most visibly as repeated ghost images of the basilar artery, the cavernous carotid segments, and the anterior cerebral artery complex propagating along the phase encoding axis and overlying adjacent parenchyma — the brainstem, the pons, the hypothalamus, or the frontal lobes depending on phase encoding direction — where they may simulate focal signal abnormalities, white matter lesions, or mass lesions if not recognised. The artefact is characteristically periodic and equidistant, which distinguishes it from true parenchymal pathology, and its displacement direction changes predictably when the phase encoding direction is swapped — a practical diagnostic manoeuvre to confirm its artefactual nature. TSE sequences are particularly susceptible because the long echo train accumulates phase errors across many echoes, and the high SNR of TSE amplifies the ghost intensity relative to conventional spin echo. Mitigation strategies include cardiac gating to synchronise acquisition with the cardiac cycle and eliminate inter-TR phase variability, spatial presaturation bands placed over the neck vessels to suppress inflowing arterial signal before it enters the imaging volume, and swapping the phase encoding direction to redirect ghost propagation away from critical anatomical regions. A fundamentally different and highly effective approach is the substitution of 2D TSE with a 3D gradient-echo inversion recovery sequence such as MPRAGE, TFE-IR, or equivalent vendor implementations: because these sequences acquire the entire brain volume within a single continuous 3D k-space readout rather than as independent 2D slices, the phase errors introduced by vascular pulsation are distributed across the full 3D k-space and their ghosting contribution to the final image is substantially reduced, making 3D acquisitions inherently more robust to pulsatile vascular artefact than their 2D TSE counterparts. For sequence-level protocol optimisation, vendor terminology and artefact management, see the dedicated MRIninja page Conventional Spin Echo (SE) Sequence.

6. Contrast Use Principles Specific to Brain MRI

Generic GBCA safety, renal screening, NSF risk and consent workflow are covered in the general MRIninja preparation and contrast pages. This section addresses brain-specific contrast decision-making only.

6.1 Non-Contrast Standard Protocol — Sufficient For

• First-line evaluation of headache without red flags and without prior imaging suggesting a mass.
• First unprovoked seizure without clinical features of neoplasia or infection.
• White matter disease monitoring in established MS between clinically stable intervals when enhancement will not change management.
• Dementia screening and morphological follow-up.
• Post-traumatic brain MRI in uncomplicated minor head injury.
• Non-focal neurological symptom workup without prior suspicious imaging.
• Post-stroke follow-up beyond the acute phase.

6.2 Gadolinium Indicated — Brain-Specific Contexts

• Suspected intracranial neoplasia: non-contrast protocol is insufficient for disease characterisation, grading and treatment planning.
• Suspected infectious or inflammatory pathology: abscess, encephalitis and meningitis may require enhancement pattern analysis.
• Active demyelination assessment: lesion enhancement defines activity when this will change management.
• Post-surgical assessment: contrast helps distinguish residual or recurrent tumour from post-surgical change.
• Cranial nerve and skull base pathology: cranial nerve courses, internal auditory canal and jugular foramen assessment usually require enhancement.
• Suspected leptomeningeal disease: delayed post-contrast 3D MPRAGE is preferred.

6.3 Post-contrast acquisition timing

The interval between gadolinium-based contrast agent injection and the start of the post-contrast T1 acquisition is not an arbitrary administrative detail but a direct determinant of diagnostic sensitivity and specificity, and must be actively managed rather than left to circumstance. Following intravenous injection, GBCA distributes rapidly into the intravascular compartment and then progressively extravasates into tissues where the blood-brain barrier is disrupted; the enhancement pattern visible on T1-weighted images reflects the local concentration of gadolinium at the time of acquisition, which evolves continuously as the contrast agent distributes, accumulates in pathological tissue, and is cleared by renal excretion.

For the majority of standard brain indications — intraparenchymal tumours, cerebral metastases, abscesses, active demyelinating plaques, and vascular malformations — the optimal acquisition window is 3 to 5 minutes after injection. Within this interval, gadolinium has achieved sufficient tissue accumulation in areas of blood-brain barrier disruption to produce diagnostically adequate T1 shortening and signal increase, while intravascular signal has partially equilibrated, reducing background vascular noise. Acquiring too early — within the first 60 to 90 seconds after injection — risks imaging during the purely intravascular phase, when parenchymal enhancement has not yet reached its maximum and small lesions may not yet be detectable. Acquiring excessively late — beyond 15 to 20 minutes for standard indications — risks contrast washout from rapidly enhancing lesions and progressive dilution of the gadolinium concentration in tissue, reducing lesion-to-background contrast.

For leptomeningeal and pachymeningeal indications — suspected leptomeningeal carcinomatosis, meningitis, neurosarcoidosis, or dural metastases — a delayed acquisition at 10 to 20 minutes after injection is recommended and substantially increases diagnostic sensitivity. The biological basis for this delay is that leptomeningeal enhancement depends on slow accumulation of gadolinium in the subarachnoid space and along the pial surface, where the volume of distribution is small and the contrast-to-noise ratio at early time points is insufficient to distinguish pathological from physiological vascular blush. At 10 to 20 minutes, gadolinium has had adequate time to accumulate in pathologically permeable leptomeningeal vessels while clearing from normal background parenchyma, maximising the signal difference between diseased and normal meninges. In this context, a 3D post-contrast FLAIR sequence acquired at the same delayed time point provides complementary sensitivity, as gadolinium in the subarachnoid space shortens the T1 of CSF and prevents its suppression by the inversion recovery pulse, rendering leptomeningeal disease visible as subarachnoid hyperintensity on FLAIR — a finding invisible on standard T1-weighted images alone.

In practical workflow terms, the injection time must be formally documented — either in the MRI system log, the acquisition notes, or the radiology report — so that the interpreting radiologist can contextualise the enhancement pattern relative to the actual timing of imaging. A post-contrast T1 acquired at 2 minutes and one acquired at 15 minutes from the same patient may show substantially different enhancement patterns for the same lesion, and without documented timing this variability cannot be interpreted or compared across serial examinations. In departments using automated power injectors integrated with the scanner console, injection time logging can be automated; where this is not available, manual documentation by the MRI technologist is mandatory.

6.4 GBCA Choice

Macrocyclic Group II agents such as gadobutrol, gadoterate meglumine and gadoteridol are the current clinical standard per ESMRMB-GREC and ESUR guidance [13]. Linear agents have been restricted or suspended by EMA for CNS indications because of gadolinium deposition evidence. Linear agent use in brain MRI should be avoided unless a macrocyclic alternative is unavailable.

Standard dose is 0.1 mmol/kg. Half-dose protocols are under investigation for selected indications with modern high-field imaging and AI-enhanced reconstruction. Double dose is no longer standard practice for most indications when high-resolution 3T MPRAGE and macrocyclic agents are available.

7. Reporting Essentials

7.1 Interpretation Framework

Brain MRI interpretation is structured around primary differential axes applied to every lesion: vascular versus inflammatory/demyelinating versus neoplastic; acute versus chronic; focal versus diffuse; intra-axial versus extra-axial; and enhancing versus non-enhancing when contrast is administered.

7.2 Mandatory Reporting Checklist

Every standard brain MRI report should directly answer the clinical question while documenting technical limitations and key negative findings when relevant. Omission of the domains below may produce an incomplete report.

• Acute or subacute DWI restriction: location, extent, vascular territory and ADC confirmation.
• Haemorrhage: location, estimated volume, pattern, surrounding oedema and mass effect.
• Obstructive hydrocephalus: site of obstruction, degree of dilatation and urgency.
• Herniation: transtentorial, subfalcine or tonsillar type and degree.
• Acute venous thrombosis: involved sinus or veins and associated DWI/SWI findings.
• White matter burden: Fazekas grade where useful, distribution pattern and atypical features.
• Infarcts: acute, subacute or chronic; territory; lacunar versus territorial; cortical versus subcortical.
• Microhaemorrhages: count or category and distribution; note cortical superficial siderosis.
• Atrophy: global versus focal; hippocampal or cortical predominance when relevant.
• Mass lesions: location, sequence signal, enhancement if contrast is given, vasogenic oedema and mass effect.
• Extra-axial and CSF spaces: subdural collections, extra-axial masses, ventricular size and hydrocephalus pattern.
• Vascular structures: dural sinus signal, arterial flow voids, visible aneurysms or vascular malformations.
• Incidental non-neurological findings: clinically relevant sinus, mastoid, orbital, scalp or skull findings.

7.3 Structured Reporting

The ESR Structured Reporting Update (2023) [14] recommends: indication, technique, systematic findings, comparison with prior studies and impression. The impression should be concise, clinically actionable and directly address the referring question.

Critical finding communication: unexpected acute findings must be communicated verbally to the referring clinician and documented in the report with time and recipient named. Brain MRI critical results include acute large infarct, haemorrhage, hydrocephalus, herniation and any new aggressive-appearing mass.

7.4 Age-Related Versus Pathological White Matter Hyperintensities

White matter hyperintensities are detected in a large proportion of patients depending on age and reporting threshold [3,12]. Fazekas grading provides standardised communication.

7.5 Incidental Findings — Clinical Decision Framework

In a series of 16,400 research brain MRI examinations, approximately 1–3% had findings requiring clinical referral [3]. Common benign or low-concern incidental variants include pineal cysts below 10 mm without hydrocephalus, arachnoid cysts in non-eloquent areas, choroid plexus cysts in adults, prominent perivascular spaces, mild age-appropriate cortical atrophy and developmental venous anomalies. Findings requiring systematic follow-up or referral include unruptured aneurysm, cavernous malformation, unexplained enhancing lesion, lobar microbleeds in a young patient and asymmetric hippocampal atrophy in a symptomatic patient.

8. MRI Technologist Pearls

8.1 Sequence Order Logic

Start with DWI. This maximises diagnostic value per unit time regardless of whether the patient can complete the full protocol. Recommended order for standard brain MRI is DWI + ADC first, sagittal T1, axial T2, axial FLAIR, SWI and then 3D T1/MPRAGE when required.

8.2 Positioning Tricks

• Foam padding bilaterally in the coil cavity centres the head and reduces rocking motion.
• A small towel roll under the neck supports the natural neck curve and reduces discomfort-driven motion.
• Head position: ears equidistant from coil sides, nose pointing directly to ceiling, no lateral rotation.
• Never start SWI or 3D FLAIR immediately after repositioning an anxious patient; wait briefly for the patient to settle.
• Use intercom check-in between sequences to maintain cooperation.

8.3 Fast Salvage Protocol

Total is about 6-7 minutes. SWI can be attempted if the patient resettles; T1 structural and post-contrast sequences can be deferred when necessary.

8.4 Common Avoidable Errors

• DWI planned on AC-PC instead of glabella–foramen magnum. This can increase anterior temporal and inferior frontal susceptibility artefact and may miss small cortical infarct.
• FLAIR TI not adjusted at 3T. Incomplete CSF suppression may simulate posterior fossa or subarachnoid pathology.
• Hearing aid or dental prosthesis not removed. These generate susceptibility artefact over the temporal bones, skull base and petrous pyramids.
• Pre-contrast T1 not acquired before injection. Baseline comparison becomes impossible and enhancement evaluation is unreliable.

11. Evidence Gaps & Ongoing Debate

12. Evidence-Based References

A. Guidelines / Consensus / Society Recommendations

B. Systematic Reviews / Meta-Analyses

C. Important Original Studies

D. Technical MRI Papers

E. Landmark Historical References