Yes—an MRI can show a brain bleed, but the answer depends on the type of bleed, its age, and the specific MRI sequences used. A brain hemorrhage (intracranial hemorrhage) is blood that has leaked from a vessel into brain tissue or the spaces around it. MRI detects it because blood contains hemoglobin, and hemoglobin's magnetic properties change as the blood breaks down over time. Each stage produces a distinct signal pattern, and a trained radiologist reads those patterns to identify and date the bleed. The following sections explain the physics, the timing, the edge cases, and the limits.
How MRI detects blood: the hemoglobin signal mechanism
MRI does not photograph tissue directly. It exploits the magnetic behavior of hydrogen protons in water and fat. When those protons sit in a strong magnetic field and absorb radio-frequency pulses, they relax back to equilibrium at rates described by two time constants: T1 (spin-lattice, longitudinal) and T2 (spin-spin, transverse). Different tissues relax at different rates, which produces image contrast.
Hemoglobin changes everything. Oxygenated hemoglobin is diamagnetic—it slightly weakens local magnetic fields. Deoxygenated hemoglobin and its breakdown products, however, are paramagnetic, meaning they carry unpaired electrons that strengthen and distort the local field. This distortion is called the susceptibility effect. It shortens T2 and T2* relaxation, so blood products appear dark on T2-weighted and susceptibility-sensitive sequences. The more concentrated the paramagnetic breakdown product, the darker the signal and the more conspicuous the bleed.
Blood also evolves chemically after it leaks out. The sequence runs: oxyhemoglobin → deoxyhemoglobin → intracellular methemoglobin → extracellular methemoglobin → hemosiderin. Each stage has its own paramagnetic profile, so each stage looks different on MRI. A radiologist can therefore infer not just that blood is present, but roughly how old it is. That dating matters clinically. A fresh bleed drives acute management decisions, while a hemosiderin rim suggests an old, resolved event.
Two specialized sequences amplify the effect. Susceptibility-weighted imaging (SWI) deliberately enhances paramagnetic contrast and detects even tiny amounts of hemosiderin or deoxyhemoglobin. Gradient-recalled echo (GRE) works on the same principle with older technology. Both are far more sensitive to small bleeds than standard T1- and T2-weighted images. This is why a protocol matters: the same patient can look normal on a routine scan and abnormal on SWI.
Quantified context: sensitivity by bleed type, size, and sequence
MRI's ability to show a bleed is not one number. It varies by hemorrhage subtype, blood volume, and sequence. Here is how the major categories compare.
| Hemorrhage type | MRI detectability | Best sequences | Key caveat |
|---|---|---|---|
| Acute intracerebral hemorrhage (within hours) | Good to excellent | SWI, GRE, T2, FLAIR | Early hyperacute (first hours) can be subtle on T1/T2 |
| Subarachnoid hemorrhage | Fair to good; depends on volume and timing | FLAIR, SWI, T2* | Small-volume SAH can be missed; CT is often first-line acutely |
| Subdural hematoma | Good across most ages | T1, T2, FLAIR | Signal changes with hematoma age |
| Epidural hematoma | Good | T1, T2, FLAIR | Usually large and conspicuous |
| Intraventricular hemorrhage | Good | FLAIR, SWI, T2* | Fluid levels and layering help |
| Microbleeds / cerebral amyloid angiopathy | Excellent with SWI | SWI, GRE | Invisible on routine T1/T2 |
| Chronic microhemorrhages (hemosiderin) | Excellent with SWI | SWI, GRE | Blooming artifact exaggerates apparent size |
The pattern is clear: sequence choice drives detection. A standard T1/T2 protocol may miss small or early bleeds that SWI captures easily. Conversely, SWI's "blooming artifact"—where the susceptibility effect makes a small deposit look larger than it is—can overstate size, so it is not a substitute for measuring the lesion on anatomical images. Blooming is the counterintuitive detail many thin explainers skip: the most sensitive sequence also distorts apparent dimensions.
Size thresholds are also approximate and sequence-dependent. Tiny petechial hemorrhages of 1–2 mm may appear only on SWI. Larger collections of several millimeters or more typically appear on conventional sequences as well. Because MRI resolution limits are in the sub-millimeter range for high-field scanners but practical detection depends on contrast, no single universal cutoff applies. The honest statement is that sensitivity rises with field strength, thinner slices, and susceptibility sequences—and falls for very small, very early, or very peripheral bleeds.
The time course: how a bleed's appearance shifts with age
Blood does not look the same on day one and day thirty. The table below summarizes the classic staging pattern for intraparenchymal hemorrhage, reconciling the standard teaching with its real-world variability.
| Stage | Approximate timing | T1 signal | T2 signal | Dominant blood product |
|---|---|---|---|---|
| Hyperacute | < 24 hours | Isointense | Hyperintense | Oxyhemoglobin |
| Acute | ~1–3 days | Isointense to slightly hypointense | Hypointense | Deoxyhemoglobin |
| Early subacute | ~3–7 days | Hyperintense | Hypointense | Intracellular methemoglobin |
| Late subacute | ~1–2 weeks | Hyperintense | Hyperintense | Extracellular methemoglobin |
| Chronic | Weeks to months | Hypointense | Hypointense (with rim) | Hemosiderin |
These ranges are teaching approximations, not hard biological boundaries. Patients vary, and concurrent conditions—anticoagulation, anemia, renal disease—shift the timetable. Anemic patients may show different signal because hemoglobin concentration is lower. Anticoagulated patients may re-bleed, layering new blood over old. This is the professional caveat worth stating plainly: the staging table is a guide to interpretation, not a clock. Radiologists weigh signal pattern against clinical history rather than reading a date off the image.
Timing also affects the choice of test. In the first hours after a suspected hemorrhage, CT remains the fastest and most widely available modality and is highly sensitive for acute blood. MRI's advantage grows for subacute and chronic bleeds, for small lesions, and for characterizing the cause—such as an underlying tumor, vascular malformation, or amyloid angiopathy. Neither test is universally superior; each answers a different question well.
Red flags and edge cases that change interpretation
Several findings should prompt urgent human review rather than reassurance from a "normal" report. The bullets below are warning signs in a clinical context, not a self-diagnosis checklist.
A sudden, severe headache ("worst of life"), new focal weakness, or a declining level of consciousness warrants immediate emergency evaluation—do not wait for outpatient imaging. A bleed seen on imaging in a patient on anticoagulants or antiplatelet drugs needs urgent reassessment, because expansion risk is elevated. Multiple lobar microbleeds on SWI raise concern for cerebral amyloid angiopathy, especially in older adults. A hemorrhage with an associated mass, edema, or midline shift suggests a structural cause and demands urgent neurosurgical input. Suspected subarachnoid hemorrhage with a negative CT but persistent symptoms may still require lumbar puncture per local protocol. New bleeding in a patient with known cancer or a vascular malformation changes the diagnostic pathway.
One edge case deserves emphasis. Subarachnoid hemorrhage—bleeding into the fluid-filled space around the brain—is the classic "MRI might miss it" scenario when performed too early or without the right sequences. FLAIR and SWI improve detection, but CT angiography and clinical judgment remain central in acute evaluation. A negative MRI does not by itself exclude a small or early subarachnoid bleed.
Common misconceptions about MRI and brain bleeds
Misconception one: "MRI always beats CT for bleeding." False. For acute hemorrhage in the emergency setting, CT is faster, cheaper, and highly sensitive, and it is the usual first-line test. MRI's edge appears in subacute and chronic bleeds, small lesions, and cause characterization. The right test depends on the clinical question and the clock.
Misconception two: "A normal MRI rules out a bleed." No imaging test is perfectly sensitive. Very small or very early bleeds can be missed, particularly on routine sequences. Microbleeds require SWI to be seen at all. A normal scan lowers probability; it does not eliminate it, and persistent symptoms warrant further evaluation.
Misconception three: "The color of a bleed on MRI tells you exactly how old it is." Signal patterns narrow the age range but overlap. Concomitant disease and re-bleeding blur the timetable. Dating is probabilistic, not exact.
Misconception four: "SWI is a stand-alone diagnostic." SWI is highly sensitive but prone to blooming artifact and can overcall small deposits. It is interpreted alongside anatomical sequences and clinical context, not in isolation.
Practical actions: what to do and when to escalate
These steps are general guidance, not a treatment plan for any individual.
For sudden severe headache, focal neurologic deficit, or altered consciousness, call emergency services immediately; do not drive yourself or wait for scheduled imaging. If a clinician orders imaging for suspected acute hemorrhage, expect CT first in most emergency settings; ask whether MRI is being added for characterization. If you have a known bleeding disorder or take anticoagulants, tell the imaging team and the referring clinician before the scan. If a report says "no acute hemorrhage" but symptoms persist, ask whether susceptibility sequences (SWI/GRE) were included and whether follow-up is warranted. If microbleeds are reported, ask about the pattern (lobar versus deep) and whether it changes management, since pattern carries diagnostic weight. Keep prior imaging available for comparison; interval change is often more informative than a single study.
Thresholds to reconcile explicitly: the timing cutoffs in the staging table (roughly 24 hours, 3 days, 7 days, 2 weeks) are conventional teaching ranges, not laboratory-validated boundaries. Size cutoffs for "small" versus "large" bleeds are similarly approximate and sequence-dependent. When a clinician cites a specific threshold, ask which sequence and which study it comes from. Different sources use different cutoffs, and the discrepancy is real, not a sign of error.
Limitations: what MRI cannot settle on its own
MRI is powerful but bounded. It cannot reliably distinguish a tiny hemorrhage from a susceptibility artifact in every case, and blooming can exaggerate apparent size. Motion artifact degrades image quality, particularly in agitated or uncooperative patients. Metallic implants and some medical devices may limit scanning or require safety screening. Claustrophobia and scan duration remain practical barriers.
Field strength matters. Higher-field scanners generally improve spatial resolution and susceptibility contrast, but availability varies and cost rises. Access is also uneven: not every facility offers the same sequences or the same radiologic expertise, so the same patient may receive different information at different centers.
Most importantly, imaging is one input among many. A radiologist's read combines the scan, the clinical history, the neurologic exam, and prior studies. An MRI report saying "hemorrhage" or "no hemorrhage" is a piece of evidence, not a verdict. The definitive interpretation belongs to the treating clinician who integrates all of it. Patients should treat imaging results as a conversation starter with their care team, not a final answer—and should seek urgent care for red-flag symptoms regardless of what any prior scan showed.