An ischemic stroke, which accounts for the vast majority of strokes, appears on MRI most characteristically as a bright white area on a sequence called diffusion-weighted imaging (DWI). That bright signal represents brain tissue where water molecules can no longer move freely because cells are swelling and dying from lack of blood flow. MRI is considerably more sensitive than CT for detecting this kind of damage, but the picture changes depending on the type of stroke, which MRI sequences are used, and how much time has passed since symptoms began.
The Bright Spot on Diffusion-Weighted Imaging
DWI is the workhorse of stroke MRI. Within minutes of a blood vessel blockage, brain cells in the affected region start to swell as their energy-dependent pumps fail. Water shifts from outside cells to inside them, and this restricted movement of water molecules shows up as a conspicuously bright region on DWI. In a landmark comparison study, the acute stroke lesion was identified correctly on DWI in every case, whereas CT picked it up in only about 42 to 63 percent of patients.1PubMed. Comparison of diffusion-weighted MRI and CT in acute stroke That gap in sensitivity is why MRI with DWI has become the preferred imaging tool when stroke is suspected and a scanner is available quickly enough.
The brightness on DWI is not the whole story, though. Radiologists also look at a computed map called the ADC (apparent diffusion coefficient) map, which essentially inverts the contrast. On the ADC map, the same stroke region looks dark rather than bright. Checking both images together helps confirm that the bright spot on DWI is genuinely restricted diffusion from acute injury and not an artifact caused by other tissue properties.
Other MRI Sequences and What They Add
A stroke MRI protocol does not rely on DWI alone. Several additional sequences each contribute a different piece of information, and together they give clinicians a much fuller picture than any single image.
FLAIR (fluid-attenuated inversion recovery) suppresses the bright signal of cerebrospinal fluid so that abnormal tissue stands out against the dark background of the brain’s fluid spaces. In acute ischemic stroke, FLAIR may initially look normal even when DWI is already bright. That delay is clinically useful: when DWI shows a lesion but FLAIR does not yet, clinicians can infer that the stroke is likely still in its early hours, a concept discussed further below.
T2-weighted images highlight areas with increased water content and are helpful in later stages of stroke when swelling and tissue breakdown are more advanced. T1-weighted images provide good anatomical detail and become useful when looking for certain types of bleeding or chronic changes. Susceptibility-weighted imaging (SWI) is particularly sensitive to blood products and can detect acute hemorrhage within six hours of symptom onset by exploiting the magnetic properties of deoxygenated hemoglobin.2PubMed. Contribution of susceptibility-weighted imaging to acute stroke assessment SWI can also reveal old microbleeds that are invisible on other sequences.
Perfusion-weighted imaging (PWI) maps blood flow through the brain in real time, typically using an injected contrast agent. In acute stroke, PWI shows the full territory of reduced blood flow, which is often larger than the DWI lesion. An alternative technique called arterial spin labeling can measure blood flow without any contrast injection at all, using magnetically labeled blood water as a natural tracer.3PubMed Central. Quantification of Serial Cerebral Blood Flow in Acute Stroke Using Arterial Spin Labeling
MRI Sensitivity Compared to CT
MRI is generally more sensitive than CT for detecting ischemia, and this advantage is especially pronounced in the first several hours after symptom onset.4PubMed Central. Comparison of CT and MR imaging in ischemic stroke CT remains the first-line scan in many emergency departments because it is faster to acquire, more widely available, and very reliable at ruling out hemorrhage. But when a CT looks normal yet stroke symptoms are obvious, MRI with DWI often reveals the culprit lesion that CT missed. For large strokes involving a major artery territory, the sensitivity gap narrows somewhat because CT can pick up early swelling and loss of tissue contrast. The gap is widest for small strokes, strokes in the brainstem or cerebellum, and very early presentations.
How the Picture Changes Over Time
The MRI appearance of a stroke is not static. It evolves through recognizable stages, and an experienced radiologist can roughly estimate the age of a stroke based on which sequences show abnormality and how the signal looks.
In the hyperacute phase (first few hours), DWI is typically the only sequence that looks clearly abnormal. FLAIR and T2 images may still appear normal or show only subtle changes. Over the next several hours and into the first day or two, FLAIR and T2 begin to light up as tissue swelling increases. By the subacute phase (days to a couple of weeks), the DWI signal can start to fade or even “pseudonormalize,” meaning the bright spot dims even though the tissue is still dead. This is a known pitfall: a pseudonormalized DWI lesion can look deceptively reassuring.
In the chronic phase, beyond roughly three weeks, the dead brain tissue softens and is replaced by fluid-filled cavities. At this stage, the area appears dark on most sequences but bright on T2-weighted images and on the ADC map, essentially the reverse of what it looked like acutely.5PubMed Central. Non-Contrast MRI Sequences for Ischemic Stroke: A Concise Overview for Clinical Radiologists These chronic cavities are a permanent scar visible on any future brain scan.
Identifying Salvageable Brain Tissue
One of the most clinically important things MRI can do in acute stroke is distinguish dead tissue from tissue that is struggling but not yet dead. The concept is called the perfusion-diffusion mismatch. The DWI lesion represents the stroke’s core, the tissue that has already suffered irreversible damage. The perfusion deficit shown by PWI represents the entire region with reduced blood flow. The difference between the two, the mismatch zone, approximates the ischemic penumbra: brain tissue at risk but potentially rescuable if blood flow is restored quickly.6PubMed Central. Magnetic resonance diffusion-perfusion mismatch in acute ischemic stroke: An update
When there is a large mismatch, a lot of brain is still salvageable, and this information can tip treatment decisions in favor of aggressive intervention like clot retrieval, even in patients who arrive hours after symptom onset. When the DWI core has already grown to match the perfusion deficit, there is little tissue left to save, and the risk-benefit calculation shifts.7PubMed Central. Perfusion-diffusion Mismatch Predicts Early Neurological Deterioration in Anterior Circulation Infarction without Thrombolysis This mismatch-based approach has reshaped how emergency teams decide who gets treated and when.
FLAIR sequences can add another clue to blood flow status. Bright-appearing blood vessels on FLAIR, called hyperintense vessels, indicate slow flow through collateral channels around the blocked area. Patients with extensive hyperintense vessels on FLAIR tend to have larger perfusion deficits and more severe hypoperfusion.8PubMed Central. Hyperintense Vessels on FLAIR: Hemodynamic Correlates and Response to Thrombolysis
When MRI Misses an Acute Stroke
DWI is highly sensitive, but it is not perfect, particularly in the very earliest minutes after a vessel occlusion. Case reports have documented hyperacute ischemic strokes that were initially invisible on DWI; the abnormality only appeared on follow-up scans done hours later. In those cases, MR angiography (which images the blood vessels) and perfusion imaging revealed the problem that DWI had not yet picked up.9Europe PMC. Hyperacute ischemic stroke missed by diffusion-weighted imaging This is rare, but it matters: a normal DWI in a patient with strong clinical signs of stroke should not be taken as a guarantee that no stroke is occurring. Repeat imaging or additional sequences can catch what DWI initially missed.
Small brainstem strokes are another known blind spot. The brainstem is physically small and surrounded by bone and flowing cerebrospinal fluid, all of which can degrade image quality. A tiny infarct there may fall below the resolution of a standard DWI scan, even though the symptoms it causes can be dramatic.
Using MRI to Estimate When a Stroke Started
Many people wake up with stroke symptoms, and nobody knows exactly when the event began. Treatment with clot-dissolving drugs has traditionally required a known onset time within a specific window, so figuring out stroke timing from imaging has real therapeutic stakes. The DWI-FLAIR mismatch is the main tool for this. When DWI shows a bright lesion but FLAIR has not yet caught up, the stroke is likely still in its early hours. In one study, about 63 percent of cases showed this mismatch within two hours of symptom onset, and the pattern faded as more time passed.10PubMed Central. Use of DWI-FLAIR Mismatch to Estimate the Onset Time in Wake-Up Strokes
Research has suggested that patients with wake-up strokes who show DWI-FLAIR mismatch may be reasonable candidates for thrombolysis, because the mismatch implies their stroke is still within a treatable window despite the unknown onset time.11PubMed. Diffusion-weighted image and fluid-attenuated inversion recovery image mismatch: unclear-onset versus clear-onset stroke This imaging-based clock has expanded treatment access for people who would otherwise have been excluded solely because they could not pinpoint when symptoms began.
Hemorrhagic Stroke on MRI
Bleeding in the brain looks completely different from ischemic stroke on MRI, and its appearance changes in a predictable sequence as the blood breaks down chemically. A hematoma passes through five distinct stages, each driven by the form of hemoglobin present. In the hyperacute phase (the first few hours), the blood contains oxyhemoglobin and tends to look bright on T2-weighted images and relatively unremarkable on T1. Within the first day or so, as oxygen is consumed and deoxyhemoglobin forms, the hematoma turns dark on T2. Over the next several days (early subacute), the hemoglobin converts to methemoglobin, first inside intact red blood cells and then free in the fluid after the cells break apart. This produces a bright signal on T1, which is the classic MRI hallmark of subacute blood. In the chronic phase, the remnants (ferritin and hemosiderin) leave a dark rim on T2-based sequences that can persist for years.12PubMed. MR appearance of hemorrhage in the brain
SWI is especially valuable for hemorrhage detection because it picks up even tiny amounts of blood products. It can identify acute hemorrhage within six hours, and it also spots old microbleeds, small deposits of hemosiderin from prior bleeding events that are invisible on standard sequences.2PubMed. Contribution of susceptibility-weighted imaging to acute stroke assessment
Hemorrhagic Transformation of an Ischemic Stroke
Sometimes an ischemic stroke converts into a hemorrhagic one. When the blood-brain barrier breaks down in the injured area and blood flow is restored (either naturally or through treatment), blood can leak into the already-damaged tissue. On MRI, this shows up as new areas of hemorrhagic signal within or adjacent to the ischemic lesion.13PubMed Central. Hemorrhagic Transformation After Ischemic Stroke: Mechanisms and Management Hemorrhagic transformation ranges from small, clinically silent petechial spots to large, symptomatic bleeds. Serial MRI is the most reliable way to detect it, and SWI sequences are particularly sensitive to the earliest signs.
What TIAs and Minor Strokes Look Like
A transient ischemic attack (TIA) has traditionally been defined by symptoms that resolve completely, but MRI has challenged that clean distinction. A pooled analysis of over 9,000 patients diagnosed with TIA found that about a third had an acute DWI lesion visible on imaging, meaning brain tissue had been permanently damaged despite symptoms resolving.14PubMed Central. Diffusion-Weighted Imaging and Diagnosis of Transient Ischemic Attack Among patients classified as having minor strokes (those with mild, lingering symptoms), the rate of DWI-positive lesions was higher, around 40 percent in one population-based study.15PubMed Central. Prognostic value of “tissue-based” definitions of TIA and minor stroke
The lesions from TIAs and minor strokes tend to be small, sometimes just a few millimeters across. They can be easy to overlook on a busy scan, but their presence matters because a DWI-positive TIA carries a higher risk of a full-blown stroke in the near future. This is one reason neurologists increasingly order MRI even when symptoms have resolved by the time a patient reaches the hospital.
Stroke Mimics and Diagnostic Traps
Not every bright spot on DWI is a stroke. Several conditions can produce restricted diffusion and fool the initial read. Brain abscesses, certain tumors (particularly high-grade gliomas), and some forms of encephalitis can all cause DWI bright spots. Seizures sometimes produce transient DWI abnormalities in the cortex that resolve within days. Migraine with aura is another well-known mimic: while DWI typically looks normal during an episode, perfusion imaging may show decreased blood flow in multiple arterial territories in up to 70 percent of cases, creating a pattern that can look alarmingly like stroke on a quick glance at perfusion maps.16PubMed Central. Magnetic resonance imaging of arterial stroke mimics: a pictorial review
The distinguishing features are usually in the details. Stroke respects vascular territories, meaning the affected area maps to a known artery’s supply zone. Mimics often cross vascular boundaries or show patterns that do not match any single vessel. Clinical context also matters enormously: a 25-year-old with a history of migraines who develops transient visual changes is in a different risk category than a 70-year-old with atrial fibrillation and sudden arm weakness.
Cerebral Venous Thrombosis
Most strokes involve arteries, but blood clots in the brain’s venous drainage system cause a different and rarer form of stroke. Cerebral venous thrombosis can be hard to spot because its MRI appearance does not follow the usual arterial stroke patterns. The clot itself may be visible as an abnormal signal in one of the major venous sinuses, and the resulting brain injury often involves swelling and bleeding that crosses typical arterial boundaries. MRI, which can image the brain tissue, blood vessels, and the clot itself, is considered a reference technique for diagnosis.17PubMed Central. Imaging of Cerebral Venous Thrombosis Dedicated venography sequences (MRV) are usually added to the protocol when venous thrombosis is suspected.
Small Vessel Disease and Chronic Markers
When someone gets a brain MRI for stroke evaluation, radiologists often find signs of chronic small vessel disease that have been accumulating silently for years. These background findings tell a story about a person’s long-term vascular health and future stroke risk.
White matter hyperintensities are bright patches on FLAIR and T2 images that appear in the brain’s white matter, particularly around the ventricles and in deeper regions. They reflect chronic microvascular damage and are extremely common in older adults and people with hypertension. Lacunes, small fluid-filled cavities from prior tiny strokes, are closely associated with white matter hyperintensities. Patients with lacunes tend to have substantially greater white matter disease and are more than twice as likely to have confluent (merging) patches of damage.18PubMed Central. The association between lacunes and white matter hyperintensity features on MRI: The SMART-MR study
Cerebral microbleeds are tiny dark dots visible on SWI or other susceptibility-sensitive sequences, each one representing a small deposit of hemosiderin from a prior microscopic bleed. Their location on MRI provides diagnostic clues about the underlying cause. Hypertension-related microbleeds tend to cluster in deep brain structures like the basal ganglia and thalamus, while those caused by cerebral amyloid angiopathy (a condition where abnormal protein deposits weaken blood vessel walls) favor the cortical surface and the junction between gray and white matter.19PubMed Central. Cerebral microbleeds: Causes, clinical relevance, and imaging approach – A narrative review One study found that a strictly cortical distribution of microbleeds was strongly associated with amyloid angiopathy, with an adjusted odds ratio around 19 compared to hypertensive disease.20Scientific Reports. Differences in lobar microbleed topography in cerebral amyloid angiopathy and hypertensive arteriopathy Even the cerebellum follows this pattern: superficial cerebellar microbleeds track with amyloid angiopathy, while deep cerebellar ones track with hypertension.21PubMed Central. Cerebellar Microbleed Distribution Patterns and Cerebral Amyloid Angiopathy
These findings matter because they influence treatment decisions. A patient with extensive cortical microbleeds from amyloid angiopathy faces a higher bleeding risk from blood thinners, which may affect whether anticoagulation is prescribed after a stroke.
Motion Artifacts and Practical Challenges
MRI requires the patient to hold reasonably still for several minutes, and stroke patients are not ideal candidates for cooperation. They may be confused, agitated, paralyzed on one side, or simply in distress. In a study of over 700 stroke MRI exams, motion artifacts were present in about 7 percent. Older age and limb motor symptoms were independently associated with motion on the scan.22PubMed. Motion artifacts and image quality in stroke MRI: associated factors and impact on AI and human diagnostic accuracy The clinical impact of motion was not uniform across findings: hemorrhage detection suffered significantly, while ischemic lesion detection held up reasonably well even on degraded images. That study also found that AI-based diagnostic tools were more vulnerable to motion artifacts than experienced radiologists, a relevant consideration as automated stroke detection becomes more common.
How Neonatal and Pediatric Strokes Differ
Strokes in newborns and young children present a special imaging challenge. The immature brain has a much higher water content than an adult brain, which makes the baseline MRI appearance different and can mask early ischemic changes on conventional sequences. Ultrasound and CT, often the first imaging obtained in a sick newborn, may look normal even when a stroke has occurred. DWI can detect neonatal stroke, but the timing of signal changes follows a different course than in adults. Pseudonormalization, where the DWI bright spot fades, occurs earlier in neonates.23PubMed. Neonatal cerebral infarction diagnosed by diffusion-weighted MRI: pseudonormalization occurs early This compressed timeline means the window for catching a neonatal stroke on DWI is narrower, and a scan done just a few days after the event may already look falsely reassuring. Pediatric neurologists and radiologists have to account for this difference when interpreting neonatal brain MRIs, and they sometimes rely more heavily on other sequences and clinical context to piece together the diagnosis.