MRI Brain – Acute Ischemic Stroke Dedicated Child Protocol

1. Executive Summary

Acute ischemic stroke MRI operates under a constraint that no other MRI examination faces: its results must directly and immediately determine eligibility for time-critical reperfusion therapy. The radiologist is not reporting a finding for clinical context — they are making a binary decision that gates thrombolysis or mechanical thrombectomy. A 10-minute difference in imaging acquisition time and a missed DWI lesion or incorrectly assessed DWI-FLAIR mismatch can determine whether a patient receives life-altering treatment within a validated therapeutic window. For sequence-level protocol optimisation, vendor terminology and artefact management, see the dedicated MRIninja page FLAIR Sequence.

This is why acute ischemic stroke requires a dedicated protocol that is fundamentally different from the generic brain MRI in three ways: it is maximally time-compressed (the entire acquisition must complete in under 15 minutes), it is diagnostically ordered (DWI is acquired first, not last), and it answers specific treament-relevant questions that the generic protocol does not even ask.

MRI has a well-defined role in the acute stroke pathway, but it competes with CT for speed. CT remains the dominant first-line imaging modality in most centres because it is universally available, fast (< 5 minutes for non-contrast CT + CTA), and excludes haemorrhage. Non-contrast MRI cannot be used for haemorrhage exclusion with the same speed as CT. The dedicated stroke MRI protocol is therefore most valuable in: (i) centres with rapid MRI access and stroke-specific MRI pathways; (ii) specific clinical scenarios where CT is insufficient — DWI-FLAIR mismatch for wake-up stroke thrombolysis eligibility; posterior fossa stroke where CT has limited sensitivity; and (iii) MRI-based patient selection for late-window thrombectomy (6–24 hours from onset) when CT perfusion is unavailable.

1.1 Added Value Over the Generic Protocol

The generic brain MRI is designed for comprehensive neurological survey. It is not designed for speed, and it does not answer the acute stroke-specific clinical questions of: (1) Is there diffusion restriction confirming acute ischaemia? (2) What is the DWI-FLAIR mismatch status (for wake-up stroke thrombolysis eligibility)? (3) Is there haemorrhage? (4) Is there a penumbra (potentially salvageable tissue beyond the ischaemic core)? (5) What is the DWI-ASPECTS score?

Specifically, the acute stroke protocol:

• Elevates DWI to the first-acquired sequence and the primary diagnostic sequence
• Adds MR perfusion (DSC or ASL) for penumbra assessment in the 6–24 hour window
• Adds MRA (intracranial and/or cervical) for vessel occlusion identification
• Uses a compressed 4–5 sequence protocol designed to complete in < 15 minutes
• Removes sequences from the generic protocol that do not contribute to the acute treatment decision (post-contrast T1, standard T2 axial, standard SWI detail — these can wait for the post-acute scan)
• Interprets every sequence in the context of a specific therapeutic binary outcome

1.2 Limits of the Dedicated Protocol

• MRI cannot replace CT for haemorrhage exclusion at the speed required in most acute centres: standard MRI protocols with GRE/SWI require 3–5 minutes and are as sensitive as CT for ICH, but the total MRI table time of 10–15 minutes vs 3–5 minutes for CT+CTA means CT remains the dominant acute pathway in most centres
• DWI has significant false-negative rate in the posterior fossa within the first 24 hours (approximately 15–20% of confirmed posterior circulation infarcts are DWI-negative at < 6 hours): a negative DWI in a patient with clinical posterior circulation stroke should not be interpreted as ruling out ischaemia
• MR perfusion requires post-processing that may not be available in emergency settings and introduces a 4–8 minute additional acquisition time
• MRA of the intracranial vessels does not replace CT angiography for speed, and requires specific knowledge of its limitations (TOF MRA overestimates stenosis, requires patient cooperation for slow acquisitions)
• Cardiac and vascular sources of embolism (cardioembolic workup, carotid plaques) require dedicated vascular protocols not described here

2. Clinical Context and Pre-Test Information

2.1 Clinical Presentation Relevant to MRI

The stroke MRI protocol is requested in three distinct clinical contexts, each requiring a different protocol emphasis:

Acute stroke with known onset time (< 24 hours): the primary clinical pathway. The question is: is there DWI-confirmed acute ischaemia? Where is it? Is there haemorrhage? Is there a penumbra for thrombectomy consideration? The NIHSS score is a direct protocol modifier — patients with NIHSS ≥ 6 and suspected LVO require intracranial MRA in addition to brain parenchyma assessment.

Wake-up stroke (unknown onset time): the patient is found with neurological deficits on awakening, with the last time known well being the time of sleep onset. The MRI question is specifically the DWI-FLAIR mismatch — positive mismatch (DWI lesion visible, FLAIR lesion absent) serves as a surrogate for onset time < 4.5 hours and is the eligibility criterion for IV thrombolysis in the WAKE-UP trial protocol [3]. This is the most distinctive MRI-specific clinical question in acute stroke — it cannot be answered by CT alone.

Late-window presentation (6–24 hours from onset or last time known well): patients who present beyond the standard 6-hour window but may still be candidates for mechanical thrombectomy. The DAWN [4] and DEFUSE-3 [5] trials established that clinical-imaging mismatch (using DWI for ischaemic core assessment and perfusion for penumbra quantification) can identify candidates for thrombectomy up to 24 hours from onset. This is the primary indication for MR perfusion in the acute stroke protocol.

2.2 Pre-Test Information the Radiologist and Technologist Must Know

Time from onset / last time known well: the single most important pre-test datum. It determines whether DWI-FLAIR mismatch is the operative clinical question, whether late-window criteria apply, and how to interpret borderline DWI-FLAIR concordance findings.

NIHSS score and lateralisation: a high NIHSS (> 10) with suspected LVO warrants immediate addition of intracranial MRA to the protocol. Lateralisation is needed to direct interpretation — right hemisphere stroke affects the dominant hemisphere in 7–10% of right-handed patients; left hemisphere stroke produces aphasia in most right-handed patients.

Suspected vessel territory: clinical localisation (anterior vs. posterior circulation) determines which DWI regions to scrutinise most carefully. Posterior circulation stroke (basilar, PICA, SCA territories) is the highest-risk scenario for DWI false-negative in the acute phase.

Prior stroke and old infarcts: chronic T2/FLAIR white matter changes, old infarcts, and chronic microangiopathy are normal in the elderly stroke population. These must be distinguished from the acute infarct on DWI and FLAIR. Prior MRI is invaluable; if unavailable, the radiologist must identify the acute lesion based on DWI restriction rather than T2/FLAIR signal alone.

Current anticoagulation or thrombolysis: affects haemorrhagic transformation risk interpretation.

Cardiac history: atrial fibrillation, recent MI, and valvular disease influence the probability of cardioembolic versus atherothrombotic mechanism — relevant to the vascular imaging priorities.

Metallic implants: all standard MRI contraindications apply. Cardiac pacemakers and certain implanted devices are a practical concern in the elderly acute stroke population. Non-compatible devices require CT/CTA as the primary pathway.

2.3 Differential Diagnosis Landscape

The acute stroke protocol must be designed to discriminate:

• Haemorrhage (intracerebral haemorrhage or subarachnoid haemorrhage): excluded by SWI/GRE — the mandatory first screen
• Stroke mimic: hypoglycaemia, Todd's paralysis post-seizure, complex migraine, transient focal neurological episode, brain tumour, demyelinating lesion — DWI restriction is typically absent in most stroke mimics; the combination of clinical context + DWI is the discriminator
• Transient ischaemic attack (TIA): clinically defined (symptoms resolved); DWI may be positive in 30–50% of clinical TIA — positive DWI confirms acute ischaemia and changes the diagnostic category
• Posterior circulation stroke vs. posterior fossa haemorrhage: critical distinction; SWI/GRE and DWI together answer this
• Venous infarction (CSVT): may produce oedema and haemorrhage; gyral distribution; DWI restriction can occur; SWI shows thrombosed vein

3. Indications, Appropriateness and Imaging Pathway

3.1 When the Dedicated Protocol Is Indicated

MRI-first for acute stroke is indicated as the primary pathway in centres with rapid MRI access (scanner immediately available, < 10 minute access from emergency department) [1]:

• All patients with wake-up stroke where thrombolysis eligibility depends on DWI-FLAIR mismatch (AHA/ASA Class IIa recommendation [1])
• Patients with clinical syndrome of posterior circulation stroke where CT sensitivity is limited (brainstem, cerebellum, occipital cortex)
• Patients presenting in the 6–24 hour window where DAWN or DEFUSE-3 eligibility is being assessed and CT perfusion is unavailable
• Patients where MRI is clinically appropriate (prior ICH treated, clear contraindication to CT iodinated contrast)

The AHA/ASA 2019 guidelines [1] state that brain imaging should be completed within 20 minutes of ED arrival in at least 50% of stroke candidates for thrombolysis. MRI-first pathways can achieve this standard if scanner access is immediate.

3.2 When the Generic Master Protocol Is Sufficient

The generic brain MRI is not sufficient for acute stroke in any time-critical scenario. It is appropriate for:

• Sub-acute stroke assessment (> 24 hours from onset), where the primary questions shift to infarct characterisation, haemorrhagic transformation assessment, and aetiological workup
• Stroke follow-up at 24–72 hours when treatment decisions have already been made
• Differential diagnosis investigation for non-acute neurological presentations where stroke is one of several possibilities

3.3 When Further Sub-Specialised Protocols Are Required

• Cervical vascular assessment: dedicated cervical MRA or contrast-enhanced neck MRA for carotid/vertebral arterial disease — relevant for secondary prevention planning but not for acute decision-making
• Cardioembolic workup: cardiac MRI for thrombus, myocardial disease — scheduled as elective after the acute event
• Cerebral venous sinus thrombosis: when CSVT is suspected — add dedicated SWI phase images and MR venography
• Aortic arch assessment for cardioembolic source: requires dedicated contrast-enhanced MRA

3.4 Red Flags Modifying Urgency or Protocol

4. Dedicated Protocol Design

4.1 Protocol Delta vs the Master Protocol

4.2 Mandatory Dedicated Sequences

4.3 Conditional and Advanced Sequences

4.4 Rationale per Disease-Specific Sequence

DWI — the Diagnostic Core of the Acute Stroke Protocol

DWI must be the first acquired sequence in the acute stroke protocol, not the last as in many generic brain MRI protocols. This is the most critical workflow departure from the standard protocol. The clinical decision about thrombolysis cannot be made until DWI is available. Acquiring T1, T2, or FLAIR before DWI in an acute stroke patient wastes minutes that directly affect neurological outcome ("time is brain" — approximately 1.9 million neurons are lost per minute in large vessel stroke [7]).

The tissue contrast logic of DWI for ischaemia is discussed in the master page. The stroke-specific semiological details are: acute ischaemia produces cytotoxic oedema — sodium-potassium pump failure causes intracellular water accumulation and reduction of extracellular space, producing restriction of water diffusion that is visible as DWI hyperintensity and ADC hypointensity within minutes of onset.

DWI-ASPECTS scoring: the Alberta Stroke Program Early CT Score adapted for DWI (DWI-ASPECTS) divides the MCA territory into 10 regions and deducts one point for each region showing acute restriction. DWI-ASPECTS ≥ 6 is a standard thrombectomy inclusion criterion [1]; ASPECTS ≤ 5 indicates large established core that predicts poor outcome from reperfusion. The radiologist must score DWI-ASPECTS for all suspected MCA territory strokes.

DWI false-negative posterior fossa: the cerebellum and brainstem are the primary sites of DWI false-negative in acute ischaemia. Susceptibility artefacts from the petrous bones, skull base, and mastoid air cells at 3T produce geometric distortion and signal dropout in the posterior fossa on EPI-DWI. A patient with clinical basilar syndrome and a negative b=1000 DWI requires: (i) review of the ADC map (a small ADC hypointensity may be present even when the DWI appears equivocal); (ii) high-b DWI (b = 2000–3000); (iii) MRA of the posterior circulation; (iv) consideration of a repeat DWI at 24 hours if the clinical suspicion remains high. For sequence-level protocol optimisation, vendor terminology and artefact management, see the dedicated MRIninja page Echo Planar DWI (EPI-DWI / SE-EPI DWI) Sequence.

b=0, 500, 1000 strategy vs b=0, 1000 only: a three-b-value acquisition (b=50–100, 500, 1000) provides a better ADC map by reducing T2 shine-through contamination. However, this adds approximately 1–2 minutes to the DWI acquisition. In the acute stroke context, the minimum acceptable protocol is b=0 + b=1000, with the ADC map calculated from these two values. Additional b-values are optional in the hyperacute setting.

SWI / GRE — Haemorrhage Exclusion

The role of SWI/GRE in acute stroke is exclusively haemorrhage detection. No thrombolysis or thrombectomy can proceed without haemorrhage exclusion. SWI (or GRE T2*) is as sensitive as CT for acute intracerebral haemorrhage [6], and substantially more sensitive for small cortical microbleeds and haemosiderin that indicate prior microangiopathy or amyloid angiopathy — relevant for haemorrhagic transformation risk assessment.

In the acute stroke protocol, SWI/GRE should be acquired before or immediately after DWI — whichever is faster in the clinical workflow. The key practical rule: never report "DWI shows acute infarct" without first confirming haemorrhage has been excluded on SWI/GRE.

SWI advantage over GRE: SWI with phase images is more sensitive for small haemorrhages, microbleeds, and cortical venous thrombosis than GRE alone. At 3T with a 2-minute acquisition, SWI provides the haemorrhage exclusion needed for the acute stroke decision.

2D FLAIR — DWI-FLAIR Mismatch Assessment

The DWI-FLAIR mismatch concept is the primary stroke-specific use of FLAIR in the acute protocol. The physiological basis: acute ischaemia produces DWI restriction within minutes of onset, but FLAIR signal elevation from vasogenic oedema and protein accumulation in the extracellular space requires several hours to develop. In the first 4.5 hours from onset, most acute infarcts show DWI restriction without FLAIR signal change (DWI-FLAIR mismatch — "DWI bright, FLAIR dark"). After 4.5 hours, FLAIR signal typically becomes visible (DWI-FLAIR match).

DWI-FLAIR mismatch specificity for onset time < 4.5 hours is 0.78, with a positive predictive value of 0.83. Sensitivity is 0.62, meaning approximately 38% of infarcts within 4.5 hours of onset already show FLAIR signal — most commonly small infarcts, posterior fossa infarcts, and infarcts in T2-bright territories [8].

The WAKE-UP trial [3] established the clinical validity of this criterion: patients with DWI-FLAIR mismatch who received IV alteplase had better outcomes than placebo-treated patients, establishing DWI-FLAIR mismatch as a validated surrogate for thrombolysis eligibility in unknown-onset-time stroke. The AHA/ASA 2019 guidelines endorse this criterion as a Class IIa indication [1].

Practical assessment of DWI-FLAIR mismatch: on the 2D axial FLAIR, the acute ischaemic region should show no FLAIR signal change (mismatch) or subtle early FLAIR change (borderline). Compare the FLAIR signal in the DWI lesion location to adjacent normal brain — the threshold is whether a clear, definite FLAIR hyperintensity is present at the DWI lesion site.

Pitfalls: small infarcts (< 15 mm) may not produce detectable FLAIR signal even after 4.5 hours — making FLAIR-based timing unreliable for small lesions [8]. Posterior fossa infarcts are particularly problematic: FLAIR signal in the posterior fossa is degraded by susceptibility artefacts and CSF-suppression challenges.

MR Perfusion — Penumbra-Core Mismatch (6–24 Hour Window)

DSC (dynamic susceptibility contrast) perfusion imaging provides haemodynamic maps of cerebral blood flow, blood volume, mean transit time (MTT), and time-to-maximum (Tmax). The standard automated processing threshold for ischaemic penumbra is Tmax > 6 seconds — tissue with delayed contrast arrival indicating impaired but potentially viable perfusion. The ischaemic core is defined as DWI-restricted tissue. The mismatch between penumbra volume (Tmax > 6 s) and core volume (DWI restriction) is the clinical tool for identifying patients with salvageable tissue beyond the standard reperfusion windows.

The DEFUSE-3 trial [5] used DWI for core and perfusion for penumbra: specific mismatch criteria (penumbra ≥ 15 mL, penumbra/core ratio > 1.8, ischaemic core < 70 mL) identified patients who benefited from thrombectomy at 6–16 hours from onset. These specific thresholds were validated in that trial and are the clinical standard for late-window thrombectomy consideration.

DSC requires intravenous contrast injection, a dedicated time series acquisition (approximately 4–6 minutes), and automated post-processing. RAPID (iSchemaView) is the most widely used commercially approved software for automated mismatch calculation and is specifically referenced in the DEFUSE-3 and DAWN trial protocols.

Contraindication: if the patient will receive IV thrombolysis within minutes, DSC contrast injection should not delay it. DSC perfusion is most useful when the treatment decision is not immediately obvious (late window, large core, uncertain clinical benefit).

Intracranial TOF MRA

Time-of-flight (TOF) MRA provides a non-contrast assessment of the intracranial arteries by saturating stationary tissue and depicting flowing blood. It identifies proximal large vessel occlusions (ICA terminus, M1 MCA, M2 MCA, basilar artery, PICA) and evaluates collateral flow patterns.

In the acute stroke protocol, TOF MRA is indicated when LVO is suspected (NIHSS ≥ 6, clinical large cortical syndrome). Its limitations for acute stroke decision-making: (i) slow acquisition (3–5 minutes); (ii) overestimates stenosis and may simulate occlusion at sites of slow or turbulent flow; (iii) cannot image vessel wall (missed dissection unless dedicated fat-suppressed T1 is added). CT angiography is more sensitive for LVO detection and is faster — but when the patient is already in the MRI scanner, intracranial TOF is the appropriate first-line vascular assessment.

4.5 Dedicated Planes, FOV, Resolution and Coverage

DWI coverage: whole brain, from the vertex to the foramen magnum. Coverage must include the cerebellum and brainstem entirely. A common technical failure in EPI-DWI is insufficient inferior coverage that excludes the posterior fossa. In the acute stroke context where posterior circulation stroke is a primary consideration, the technologist must verify that the inferior extent of DWI coverage includes the medulla and the entire cerebellum on the localiser. For sequence-level protocol optimisation, vendor terminology and artefact management, see the dedicated MRIninja page Echo Planar Imaging (EPI) Sequence.

FLAIR: standard 2D axial 5 mm is sufficient for DWI-FLAIR mismatch assessment — the 3D isotropic FLAIR that is preferred for MS and epilepsy is not required for the acute stroke question and adds acquisition time. The DWI-FLAIR mismatch comparison requires only visual assessment of whether FLAIR signal change is visible at the DWI lesion location.

SWI: whole brain axial, 1.5–2 mm slice thickness. Phase images must be retained for cortical venous assessment (CSVT differential).

TOF MRA: axial acquisition covering the intracranial vessels from the carotid siphons to the distal MCA branches. MIP reconstructions in multiple planes are required for interpretation. The MRA acquisition is prescribed after the DWI is available, so the technologist knows which territory to prioritise.

MR perfusion: whole brain axial EPI, 5–7 mm slice thickness (SNR constraint for DSC), same coverage as DWI.

4.6 Contrast Strategy

Gadolinium-based contrast agents are not required for the core acute stroke MRI protocol (DWI + FLAIR + SWI + TOF MRA). This is an important advantage of MRI over CT in some scenarios — the primary acute stroke MRI decisions can be made without contrast.

Contrast is used in the acute stroke protocol specifically for:

• DSC MR perfusion in the late-window patient (6–24 hours): standard dose, power injector, 3–4 mL/s injection rate
• Contrast-enhanced MRA when better posterior circulation or cervical vessel visualisation is required

When thrombolysis is planned within minutes, gadolinium injection (if required for perfusion) should not delay treatment. Perfusion imaging is for the late-window decision where minutes are less critical than in the hyperacute window.

4.7 Sequence Matching, Reproducibility and Follow-Up

For serial stroke MRI (24-hour follow-up after thrombolysis or thrombectomy), the follow-up scan should include: DWI (for new infarct or expansion assessment), FLAIR (for established infarct visibility), SWI/GRE (for haemorrhagic transformation), and T1 post-contrast (for blood-brain barrier disruption assessment and luxury perfusion). The follow-up protocol at 24 hours is close to the generic brain MRI protocol and includes sequences that were deferred from the acute protocol.

At 24 hours, infarct size on DWI or FLAIR provides the primary outcome assessment. FLAIR infarct volume at 24 hours is the standard clinical trial endpoint for tissue outcome [1, 4, 5].

5. MRI Semiotics — Disease-Specific Imaging Findings

5.1 Direct Signs

DWI Restriction — Acute Ischaemia

Hyperacute phase (0–6 hours): DWI hyperintensity with corresponding ADC hypointensity (true diffusion restriction) appears within minutes to hours of onset. The DWI lesion may be smaller than the eventual infarct (it represents the ischaemic core — the most severely ischaemic tissue). At 0–3 hours, DWI lesion volume underestimates the final infarct volume by an average of 20–30%.

Acute phase (6–24 hours): DWI lesion becomes more conspicuous and may enlarge as ischaemia progresses. ADC values are at their nadir (lowest point) — typically 30–50% below normal white matter ADC.

Subacute phase (24 hours–2 weeks): DWI hyperintensity persists due to T2 shine-through effect even as true restriction diminishes. ADC values begin to pseudonormalise (return toward normal values) after approximately 5–10 days, then eventually become elevated (vasogenic oedema and tissue breakdown).

ADC pseudonormalisation: the ADC map returns to apparently normal values at approximately 1–2 weeks despite the lesion still being visible on DWI (due to T2 shine-through). This temporal progression is an important marker for infarct age determination.

Grading Acute Ischaemic Core: DWI-ASPECTS

The DWI-ASPECTS (Alberta Stroke Program Early CT Score on DWI) scores the MCA territory on two axial slices:

• Upper slice (basal ganglia level): caudate (C), lentiform nucleus (L), internal capsule (IC), insular ribbon (I) — 4 structures
• Lower slice (centrum semiovale level): M1–M6 (6 cortical zones of the MCA territory)

Each region with acute DWI restriction loses 1 point from a maximum of 10. Scoring:

• ASPECTS 10: no MCA territory restriction
• ASPECTS 7–9: small to moderate ischaemic core — favourable prognosis for thrombectomy
• ASPECTS ≤ 6: larger ischaemic core — increasing risk of poor outcome despite reperfusion
• ASPECTS ≤ 5: large ischaemic core — associated with poor outcome even with successful reperfusion in most studies

DWI-ASPECTS is scale-specific for the MCA territory and does not apply to posterior circulation strokes. The NIHSS ≥ 6 + DWI-ASPECTS ≥ 6 combination defines standard thrombectomy eligibility for the anterior circulation.

DWI-FLAIR Mismatch

Positive mismatch (DWI bright, FLAIR normal or no definite change): supports onset time < 4.5 hours; supports thrombolysis eligibility in wake-up stroke and unknown onset time (WAKE-UP criteria [3]).

Negative mismatch / DWI-FLAIR match (DWI bright, FLAIR also showing clear hyperintensity): supports onset time > 4.5 hours from onset; thrombolysis may be contraindicated in wake-up stroke if clear FLAIR signal is present.

Interpretation caveats:

• Small infarcts (< 15 mm) frequently show DWI-FLAIR mismatch regardless of time, because small lesions may not generate sufficient vasogenic oedema to produce detectable FLAIR signal even after 6–12 hours — this means DWI-FLAIR mismatch is unreliable for timing in small infarcts [8]
• Posterior fossa infarcts: FLAIR is unreliable in the posterior fossa due to susceptibility and CSF challenges — DWI-FLAIR mismatch assessment should be restricted to supratentorial infarcts
• Large infarcts may show FLAIR signal early (< 3 hours) because the larger oedematous volume produces earlier vasogenic oedema

Haemorrhage on SWI/GRE

Acute intracerebral haemorrhage: T2* signal loss (blooming artefact) exceeding the anatomical size of the clot. At 3T SWI, the blooming is more extensive than at 1.5T — this is an advantage (greater sensitivity for small haemorrhages) but requires awareness that the SWI-defined haematoma size overestimates the true haematoma volume.

Haemorrhagic transformation: petechial or confluent T2* signal loss within a previously ischaemic territory. The ECASS (European Cooperative Acute Stroke Study) classification of haemorrhagic transformation is used in reporting: HI1 (petechial) and HI2 (confluent haemorrhagic infarction) vs. PH1 and PH2 (parenchymal haematoma — the symptomatic forms).

5.2 Indirect and Secondary Signs

Hyperintense vessel sign (FLAIR/DWI): intravascular T2/FLAIR hyperintensity in a major vessel indicates slow flow or thrombosis. The FLAIR vascular hyperintensity (FVH) in MCA territory vessels visible as bright serpentine signal in the sylvian fissure indicates impaired antegrade flow and is an indirect sign of proximal occlusion. FVH distal to the DWI lesion (FVH-DWI mismatch) is a proposed surrogate for penumbra, though less validated than DSC perfusion.

Cortical and arterial T2/SWI hypointensity*: the "susceptibility vessel sign" — a linear T2* hypointensity within a cerebral artery (particularly M1 MCA) on SWI, corresponding to an acute thrombus (deoxyhaemoglobin content). This sign is highly specific for LVO and correlates with the dense MCA sign on CT.

Loss of flow-void in major arteries: on T2-weighted images, major cerebral arteries normally appear as flow voids. Loss of the MCA or basilar artery flow void on T2 or FLAIR suggests slow flow or occlusion.

5.3 Severity, Extent and Activity Assessment

For acute stroke, severity is assessed by:

• DWI-ASPECTS for MCA territory infarcts (as above)
• DWI infarct volume (mL): measured or estimated from DWI; core volume < 70 mL is a DEFUSE-3 eligibility criterion [5]
• Presence of penumbra: Tmax > 6 s volume on DSC perfusion; penumbra/core mismatch ratio > 1.8 defines significant salvageable tissue
• NIHSS correlation: the clinical stroke severity correlates with DWI lesion volume and location; higher NIHSS + small DWI lesion suggests LVO with surviving penumbra (high-priority thrombectomy candidate)

5.4 Validated Classification and Grading Systems

DWI-ASPECTS (Pexman et al., adapted for DWI)

As described in Section 5.1 — 10-point scale for MCA territory ischaemic core quantification on DWI. Originally designed for non-contrast CT (ASPECTS) and validated for DWI. Used in every major thrombectomy trial as a patient selection criterion [4, 5].

ECASS Classification for Haemorrhagic Transformation

• HI1: small petechiae along infarcted cortex
• HI2: confluent petechiae within infarct zone, no space-occupying effect
• PH1: haematoma ≤ 30% of infarct zone, mild space-occupying effect
• PH2: haematoma > 30% of infarct zone, significant space-occupying effect

PH1 and PH2 are symptomatic haemorrhagic transformation (worsening of ≥ 4 NIHSS points or death attributed to haematoma); HI1 and HI2 are typically asymptomatic.

5.5 Differential Diagnosis on MRI

5.6 Mimickers, Pseudolesions and Normal Variants

T2 shine-through on DWI: CSF in ventricles, large perivascular spaces, and areas of chronic high T2 signal appear bright on DWI due to T2 contamination, not true diffusion restriction. Confirmed by the ADC map — T2 shine-through shows normal or elevated ADC; true restriction shows low ADC.

Pseudorestriction in the subacute phase (ADC pseudonormalisation): around 10–14 days, the ADC returns to near-normal values while DWI remains bright. Misinterpreting this as "no diffusion restriction" leads to false assurance. The signal on b=1000 images at this stage reflects T2 shine-through, not active ischaemia. In clinical practice, the DWI lesion at 2–3 weeks should be confirmed on FLAIR, which remains abnormal well into the chronic phase.

Magic angle and posterior fossa geometry artefacts: the posterior fossa EPI distortion can simulate small DWI lesions at the cortical surface of the brainstem or cerebellum. Always confirm with the ADC map — true ischaemia shows ADC hypointensity; artefact typically shows a normal or mildly elevated ADC.

6. Reporting Framework Specific to Acute Stroke

6.1 Structured Reporting Template

Indication: state the clinical scenario (acute onset right/left hemisphere deficit; NIHSS; last time known well; suspected anterior vs. posterior circulation; wake-up stroke; late-window thrombectomy candidate).

Technique: DWI + ADC, b-values used; SWI/GRE; FLAIR; MRA (if acquired); DSC perfusion (if acquired, contrast dose and timing). State total acquisition time.

Comparison: prior brain MRI (if available) — old infarcts must be documented to distinguish from acute findings.

Findings — in treatment-relevant order:

Haemorrhage (SWI/GRE — must be stated first):

• Intracerebral haemorrhage: absent / present (location, volume)
• Haemorrhagic transformation of prior infarct: absent / present
• Microbleeds: absent / present (location, number — relevant for haemorrhagic transformation risk)

Acute ischaemia (DWI + ADC):

• DWI lesion: present / absent
• If present: location (vascular territory, side, lobe/structure), maximum dimension, DWI-ASPECTS score (for MCA territory)
• ADC confirmation: restricted (ADC hypointense) / equivocal / T2 shine-through only
• Estimated DWI lesion volume (mL) if relevant for late-window decisions

DWI-FLAIR mismatch assessment (FLAIR):

• FLAIR signal at DWI lesion location: absent (mismatch — within 4.5 hours) / equivocal / present (match — beyond 4.5 hours)
• Limitations stated if posterior fossa or small lesion applies

Vascular occlusion (MRA, if acquired):

• Large vessel occlusion: present / absent
• If present: location, side (ICA terminus, M1, M2, basilar, etc.)
• Collateral assessment: absent / poor / moderate / good

Penumbra (DSC perfusion, if acquired):

• Tmax > 6 s volume (mL) — penumbra estimate
• DWI core volume (mL) — ischaemic core estimate
• Mismatch ratio (penumbra/core)
• DEFUSE-3 eligibility criteria: met / not met (state specific criteria)

Background:

• Prior infarcts: absent / present (location, approximate age)
• White matter changes: grade (Fazekas)
• Other relevant findings

Impression — treatment-relevant:

• Is there acute ischaemia? Yes / No / Equivocal
• Is there haemorrhage? Yes / No
• DWI-FLAIR mismatch status (for wake-up stroke): Positive mismatch / Match / Not applicable / Uninterpretable (posterior fossa / small lesion)
• DWI-ASPECTS score (if MCA territory)
• LVO present/absent
• Penumbra-core mismatch: significant / not significant / not assessed
• DEFUSE-3 criteria met: yes / no / not assessed

6.2 Mandatory Disease-Specific Reporting Checklist

• Haemorrhage explicitly excluded (SWI/GRE reviewed and stated)
• DWI lesion: present or absent — stated explicitly
• ADC hypointensity confirmed (not T2 shine-through alone)
• Vascular territory identified
• DWI-ASPECTS scored (for MCA territory strokes)
• DWI-FLAIR mismatch: stated explicitly with limitation if applicable
• LVO present or absent (if MRA acquired)
• Haemorrhagic transformation: assessed (SWI)
• Prior infarcts documented (to distinguish from acute)
• Total acquisition time documented

6.3 Critical Findings and Communication

Immediate direct verbal communication to the treating physician is required for:

• Acute DWI lesion confirmed in a patient within the thrombolysis window — treatment must proceed within minutes
• LVO identified — thrombectomy team must be activated immediately
• Intracerebral haemorrhage — thrombolysis is contraindicated; haemostatic management pathway activated
• DWI-FLAIR mismatch positive in a wake-up stroke patient — thrombolysis eligibility confirmed
• Basilar artery occlusion — critical emergency; highest mortality acute stroke

Every acute stroke MRI report must reach the treating physician within 10 minutes of scan completion. Verbal confirmation followed by written report.

6.4 Common Reporting Errors

7. Technical Pitfalls and Disease-Specific Optimisation

7.1 Technical Pitfalls Specific to Acute Stroke

EPI susceptibility artefacts in the posterior fossa: the single most important technical limitation of DWI in acute stroke. EPI sequences are highly susceptible to field inhomogeneities at air-tissue interfaces — the mastoid air cells, petrous bones, and skull base produce severe geometric distortion and signal dropout in the posterior fossa at 3T. A small brainstem or cerebellar infarct in the acute phase can be invisible or distorted to the point of undetectability on standard EPI-DWI at 3T. Mitigation: use wider bandwidth for posterior fossa EPI; apply B0 distortion correction; review the ADC map even when the b=1000 image appears normal; consider high-b DWI (b = 2000–3000).

DWI false-negative at < 6 hours in posterior circulation: approximately 15–20% of confirmed posterior circulation infarcts are DWI-negative within the first 6 hours, even on technically adequate scans at 3T. This is a fundamental biological and technical limitation — cytotoxic oedema may not be established, and susceptibility artefacts in the posterior fossa reduce sensitivity. A negative DWI in a patient with basilar syndrome is not a reliable ruling-out of ischaemia. The MRA is the critical complementary sequence — basilar or posterior cerebral artery occlusion on MRA in a DWI-negative clinical posterior stroke justifies treatment on the basis of the vessel imaging alone.

Motion-induced DWI artefacts in acute stroke patients: many acute stroke patients have some degree of involuntary head or limb movement due to their neurological deficit. EPI-DWI is relatively motion-tolerant (single-shot EPI acquires the entire k-space in 50–100 ms), but head motion between EPI acquisitions of different b-values can produce ADC map registration errors. PROPELLER/BLADE DWI (motion-robust segmented readout) is an option for patients with significant motion but adds acquisition time. For sequence-level protocol optimisation, vendor terminology and artefact management, see the dedicated MRIninja page Spin Echo DWI / Non-EPI DWI Sequence.

ADC pseudonormalisation misread as non-restricted: as described in Section 5.6, the ADC "normalises" around 10–14 days even though the infarct is visible on DWI (T2 shine-through). In the acute protocol this is less relevant, but in a patient with a subacute stroke re-examined at 7–10 days, the ADC normalisation can be misread as "no evidence of ischaemia" if only ADC is reviewed without DWI.

7.2 Sequence-Specific Disease Pitfalls

FLAIR and DWI-FLAIR mismatch in the posterior fossa: as stated above, FLAIR signal in the brainstem and cerebellum is inherently difficult to assess due to CSF pulsation artefacts, susceptibility effects, and the small size of posterior fossa structures. The DWI-FLAIR mismatch concept has not been validated for posterior fossa strokes — do not apply it there.

Susceptibility vessel sign overread: the T2* hypointensity of a thrombosed MCA ("susceptibility vessel sign") can be mimicked by a normal tortuous vessel or by calcification in the vessel wall in elderly patients. Correlation with the TOF MRA (showing loss of flow signal) is required to confirm the susceptibility vessel sign as thrombus.

SWI blooming overestimation of ICH volume: at 3T, the SWI blooming effect overestimates the true haematoma volume by approximately 30–40% compared to CT. This is clinically relevant when haematoma volume is being used for management decisions (surgical evacuation threshold). State in the report that haematoma volume on MRI/SWI may overestimate true volume and CT provides more accurate volumetry.

7.3 When the Exam Is Non-Diagnostic for This Question

• DWI-negative posterior fossa stroke: if clinical suspicion remains after DWI-negative MRI, CT angiography of the posterior circulation is the recommended next step. Do not exclude posterior stroke based on MRI alone.
• Severe patient motion: if motion artefacts degrade DWI or FLAIR quality to the point of non-diagnostic imaging, CT + CTA is the appropriate fallback for the acute treatment decision.
• Metallic implant or MRI-incompatible device: CT + CTP is the primary pathway for acute stroke assessment in these patients.

8. MRI Technologist Pearls Specific to Acute Stroke

8.1 Disease-Specific Positioning and Coil Tricks

In the hyperacute stroke context, every step that delays DWI acquisition is clinically consequential. The technologist should:

• Have the stroke MRI protocol pre-set on the scanner as a single-push-button acquisition with all four sequences pre-defined in order
• Not perform pre-scan adjustments (shimming, coil calibration) that delay DWI acquisition — the scanner should be set for automatic shimming
• Immediately after patient positioning, start DWI as the first sequence without waiting for detailed localiser review

For patients with involuntary limb movement (typical in large hemisphere strokes), foam padding around the head reduces minor motion. The arm on the side of the stroke should be secured with a light strap if it is showing involuntary movements.

8.2 Sequence Order Logic

Mandatory sequence order for acute stroke MRI:

1. Three-plane localiser (30 seconds)
2. DWI ← first; no exceptions; this is the therapeutic decision sequence
3. SWI/GRE ← second; haemorrhage exclusion cannot proceed to step 4 without this
4. 2D FLAIR ← third; DWI-FLAIR mismatch assessment
5. TOF MRA ← fourth, if LVO suspected (plan from DWI to confirm territory and guide TOF coverage)
6. DSC perfusion ← last; only in late-window patients; requires contrast

Total targeted acquisition time: DWI (2 min) + SWI (3 min) + FLAIR (2–3 min) + optional TOF MRA (3–4 min) = 10–12 minutes for the haemorrhage + ischaemia + mismatch assessment.

8.3 Fast Salvage Version of the Dedicated Protocol

Three sequences in 6–8 minutes provide the minimum clinically actionable acute stroke MRI. All thrombectomy and thrombolysis eligibility questions can be answered from these three sequences when combined with the NIHSS and onset time.

8.4 Disease-Specific Avoidable Errors

9. Quality Control Checklist for the Dedicated Protocol

• DWI acquired as first sequence (not after T1, T2, or FLAIR)
• ADC map generated and reviewed
• DWI inferior coverage includes posterior fossa (medulla and cerebellum visible)
• SWI or GRE acquired and reviewed for haemorrhage
• SWI phase images saved and available
• FLAIR acquired (for DWI-FLAIR mismatch assessment)
• DWI-FLAIR mismatch explicitly assessed and documented in report
• DWI-ASPECTS scored for MCA territory strokes
• MRA acquired and MIP reconstructions available if LVO was suspected
• DSC perfusion maps (Tmax, CBF, core/penumbra volumes) processed and available if acquired
• Total scan time documented
• Verbal report communicated to treating team within 10 minutes of DWI completion

11. Evidence Gaps and Ongoing Debate

MRI vs CT as first-line imaging in acute stroke: the debate about whether MRI or CT should be the primary imaging modality for all acute stroke patients is unresolved. CT+CTA is faster and universally available; MRI provides DWI confirmation, DWI-FLAIR mismatch, and superior posterior fossa assessment. Most guidelines endorse CT as the standard initial pathway with MRI used for specific clinical questions (wake-up stroke, posterior circulation, late window).

DWI-FLAIR mismatch reliability at the margins: the moderate sensitivity (0.62) of DWI-FLAIR mismatch for the < 4.5-hour time window — driven primarily by small infarcts and posterior fossa lesions — means that approximately 38% of eligible patients may be excluded from thrombolysis by a strict FLAIR-visible sign. Whether a more liberal application of DWI-FLAIR mismatch criteria would improve outcomes or increase haemorrhagic risk is an active research question.

AI-automated DWI-FLAIR mismatch assessment: inter-observer variability in visual DWI-FLAIR mismatch assessment is moderate — the WAKE-UP trial reported κ = 0.85 for expert readers, which may not reflect non-specialist centres. Deep learning models for automated DWI-FLAIR mismatch prediction (including DWI-only models) are in development [see recent literature], but no model has been prospectively validated for clinical deployment at the standard required for thrombolysis decisions.

Perfusion imaging thresholds beyond DEFUSE-3: the Tmax > 6 s threshold for penumbra was validated in DEFUSE-3 but is not universally accepted as the optimal threshold. Different studies and clinical contexts may justify different Tmax thresholds. The standardisation of perfusion processing across vendors and algorithms remains incomplete.

Large infarct core thrombectomy (DWI-ASPECTS ≤ 5): several recent trials (ANGEL-ASPECT, SELECT2, TENSION) have challenged the established threshold of DWI-ASPECTS ≥ 6 for thrombectomy, demonstrating potential benefit even in large core infarcts (ASPECTS 3–5, DWI volume > 70 mL). This is an active and unresolved area that is changing clinical practice in some centres.

MRA vs CTA for acute LVO detection: CTA is more sensitive than TOF MRA for vessel occlusion and provides better posterior fossa assessment. In centres where MRI is first-line for acute stroke, whether TOF MRA is equivalent to CTA for treatment decision-making is not definitively established.

12. Evidence-Based References

A. Guidelines / Consensus / Society Recommendations

B. Systematic Reviews / Meta-analyses

C. Important Prospective / Original Studies

D. Technical MRI Papers

E. Landmark Historical References

End of document — MRI Brain Acute Ischemic Stroke Child Protocol — MRIninja v1.0 — May 2026 Parent page: MRI Brain Generic Standard Protocol Future child pages building on this page: TIA MRI protocol; large core stroke assessment; haemorrhagic transformation evaluation; post-thrombectomy follow-up MRI; basilar artery syndrome.