A CT scan of the brain produces cross-sectional images that reveal the physical structure inside your skull: bone, brain tissue, fluid-filled spaces, and blood. Because different tissues absorb X-rays at different rates, a CT can distinguish between a fresh bleed, swollen brain tissue, a fluid-filled cavity, a calcified mass, and normal anatomy in a matter of seconds. That speed is why it remains the go-to first imaging test in emergency rooms worldwide, even though it has real blind spots that sometimes call for other tools.
How the Image Is Built
During a head CT, an X-ray tube rotates around your head while detectors on the opposite side measure how much radiation passes through. Dense structures like bone absorb more X-rays and appear bright white on the resulting image. Air absorbs almost none and looks black. Brain tissue, cerebrospinal fluid, fat, and blood each fall somewhere in between, producing various shades of gray. The scanner’s computer assembles these measurements into thin slices, typically a few millimeters thick, that can be scrolled through from the top of the skull to the base. The whole process takes about 30 seconds of actual scanning time.
The concept dates to 1971, when the first patient CT examination was performed by Godfrey Hounsfield and James Ambrose, an achievement widely considered the greatest revolution in medical imaging since the discovery of X-rays themselves.1PubMed Central. How CT happened: the early development of medical computed tomography That first scan took hours to produce a single image of the brain. Modern scanners generate hundreds of images in under a minute, and the technology has expanded to cover the entire body.2PubMed Central. Milestones in CT: Past, Present, and Future
Bleeding Inside the Skull
Fresh blood is denser than brain tissue, so it shows up as a bright white area on a non-contrast CT. This is one of the scan’s greatest strengths and the main reason it is ordered in emergencies. A doctor can quickly see whether blood is collecting between the skull and brain (epidural or subdural hematoma), spreading across the brain’s surface (subarachnoid hemorrhage), or pooling within the brain tissue itself (intracerebral hemorrhage). The location and volume of the bleed guide decisions about whether surgery is needed, how aggressively to manage blood pressure, and what the likely prognosis looks like.
Hemorrhagic stroke, where a blood vessel bursts inside the brain, is diagnosed almost instantly on a plain CT. This matters because the treatment for a hemorrhagic stroke is the opposite of what you’d give for a clot-caused (ischemic) stroke. Giving a clot-busting drug to someone who is actively bleeding could be fatal. The CT’s ability to rule out hemorrhage within minutes is the single most time-critical piece of information in acute stroke care.
Ischemic Stroke and Its Limits
Ironically, while CT excels at spotting bleeding, it struggles with the other kind of stroke. In an ischemic stroke, a clot blocks blood flow to part of the brain, and the affected tissue starts to swell and die. In the first several hours, this swelling may be too subtle to see on a standard CT. Detecting acute ischemic stroke on CT images is genuinely challenging even for experienced radiologists. One study found that the sensitivity of detecting these early changes was only around 37% for a deep-learning model and about 33% for a board-certified radiologist reading the scan without computer assistance.3PubMed. Automatic detection of acute ischemic stroke using non-contrast computed tomography and two-stage deep learning model That means roughly two-thirds of acute ischemic strokes were missed on the initial plain CT. The scan’s main job in this scenario is less about confirming the clot and more about ruling out bleeding so that clot-busting treatment can proceed safely.
The picture gets worse in certain brain regions. The posterior fossa, which houses the brainstem and cerebellum at the back and base of the skull, is particularly hard to evaluate on CT because of bone artifact from the surrounding thick skull base. One study calculated the sensitivity of non-contrast head CT for posterior fossa strokes at roughly 42%, and brainstem strokes fared even worse at about 33%.4PubMed Central. Comparative Sensitivity of Computed Tomography vs. Magnetic Resonance Imaging for Detecting Acute Posterior Fossa Infarct A separate study in a different population reported somewhat higher sensitivity for plain CT in the posterior fossa (around 77%), though still below what MRI achieves, and with specificity around 65%.5Medical Journal of Eastern Nepal. Comparative study of plain CT head versus MRI brain for detecting acute posterior fossa infarct The takeaway is consistent across studies: if a posterior fossa stroke is suspected and the CT looks normal, MRI is typically the next step.
CT Perfusion and Angiography
To compensate for the limitations of a plain scan in stroke, hospitals often add perfusion CT and CT angiography (CTA) to the protocol. CT perfusion involves injecting contrast dye and then scanning the brain repeatedly as the dye washes through. This creates a map of blood flow, blood volume, and transit time across different brain regions. Areas getting poor blood flow light up as abnormal, revealing the tissue at risk even when the plain images look normal.
CTA takes a slightly different approach: it images the blood vessels themselves after contrast injection. This lets doctors see whether a major artery is blocked. In one study, combining four-dimensional CTA with perfusion maps pushed the sensitivity for detecting large vessel occlusions to 94%, with a positive predictive value of 100%, meaning virtually no false alarms.6PubMed. Angiographic reconstructions from whole-brain perfusion CT for the detection of large vessel occlusion in acute stroke CTA is also used outside of stroke to evaluate aneurysms and other vascular malformations in the brain, though it is not perfect. The most commonly missed findings on head and neck CTA are small aneurysms, which account for about half of all interpretation errors.7PubMed Central. Interpretation errors in CT angiography of the head and neck and the benefit of double reading
Traumatic Brain Injury
After a head injury, CT is the standard first-line test. It can show skull fractures, bleeding inside or around the brain, bruising of brain tissue (contusions), and swelling. One of the most clinically important measurements a doctor extracts from a trauma CT is midline shift, which is how far the brain’s central structures have been pushed to one side by swelling or a blood collection. The degree of midline shift is closely correlated with patient outcomes in traumatic brain injury.8PubMed Central. Midline Shift vs. Mid-Surface Shift: Correlation with Outcome of Traumatic Brain Injuries Even a few millimeters of shift can signal a surgical emergency, because it means the brain is being compressed.
CT also catches air inside the skull (pneumocephalus), which can indicate a fracture that has breached one of the sinuses. And it can identify foreign objects, bone fragments driven into the brain, or depressed skull fractures where a piece of bone has been pushed inward. For patients who are unconscious or too unstable for a longer MRI scan, the speed and availability of CT make it irreplaceable in trauma settings.
Fluid Spaces and Hydrocephalus
The brain’s ventricles are fluid-filled chambers that produce and circulate cerebrospinal fluid (CSF). On a CT scan, CSF appears dark compared to the surrounding brain tissue, making the ventricles easy to identify and measure. When the ventricles are abnormally enlarged, it may point to hydrocephalus, a condition where CSF accumulates because it is not draining properly. This buildup increases pressure inside the skull and can cause headache, vision problems, cognitive decline, and, if untreated, serious neurological damage.9PubMed. Attention-Driven Deep Learning for Hydrocephalus: Preliminary Benchmarking of Multiscale 3D Attention Against Conventional 3D CNN Models Using Ventricular Enlargement on CT
Radiologists use standardized measurements on CT to gauge how enlarged the ventricles are. The Evans index, a ratio comparing the width of the frontal horns of the ventricles to the maximum internal diameter of the skull, is one of the most common. Research has shown that the Evans index and third ventricle width are the strongest independent predictors for distinguishing hydrocephalus from other causes of enlarged ventricles, such as brain atrophy after a stroke or injury.10PubMed Central. Comparative CT Ventricular Morphometrics in Hydrocephalus, Stroke, and Traumatic Brain Injury: A Distortion-Controlled Analysis This distinction matters because the treatments are completely different: hydrocephalus may require a shunt or surgical procedure to drain the fluid, while atrophy-related enlargement usually does not.
Ventricular enlargement can also be one-sided. A case report described a patient whose left lateral ventricle was dilated due to post-stroke scarring in the left frontal and temporal lobes, which obstructed CSF pathways on that side.11PubMed Central. When Stroke Leads to Hydrocephalus: A Case of Unilateral Ventricular Enlargement CT made this asymmetry immediately visible and guided further workup.
Tumors, Masses, and Calcifications
Brain tumors often show up on CT as areas of abnormal density, sometimes with surrounding swelling that pushes adjacent structures aside. A plain CT can detect many larger tumors, and adding contrast dye usually makes them stand out more because tumors tend to have leaky blood vessels that allow the dye to seep in, creating a bright area called enhancement. CT is often the first test that raises the alarm, though MRI typically follows for more detailed characterization of the tumor’s borders and type.
Calcifications are another common finding. Certain normal brain structures, like the pineal gland and choroid plexus, calcify with age, and these are considered harmless. Some tumors, such as meningiomas and craniopharyngiomas, calcify in characteristic patterns that can help narrow the diagnosis. CT is actually better than MRI at detecting calcification because calcium is very dense and lights up conspicuously on CT.
Age-Related Changes
As you get older, the brain gradually loses volume, a process visible on CT as widened sulci (the grooves on the brain’s surface) and enlarged ventricles. This is sometimes called brain atrophy. Some degree of atrophy is normal with aging, but more severe or asymmetric atrophy can suggest a neurodegenerative disease. CT can also reveal leukoaraiosis, a term for low-density white matter changes that reflect chronic small vessel disease. Both atrophy and leukoaraiosis can be assessed visually on a plain head CT, and research suggests they may serve as surrogate markers of a person’s baseline functional status before a stroke, potentially influencing treatment decisions.12PubMed. Brain Atrophy and Leukoaraiosis Correlate with Futile Stroke Thrombectomy
These changes are often reported in the radiology read even when they weren’t the reason the scan was ordered. If you’re getting a head CT after a fall and the report mentions “age-appropriate atrophy” or “mild small vessel ischemic changes,” that’s generally describing the background wear and tear of an aging brain, not a new problem requiring treatment.
Incidental Findings
One thing that surprises many people is how often a brain CT turns up something unexpected. In a study of head injury patients at a Nigerian hospital, incidental findings were noted in about 85% of scans. The most common was intracranial calcification, present in roughly 62% of patients, and over 90% of all incidental findings were benign.13PubMed Central. Incidental cranial CT findings in head injury patients in a Nigerian tertiary hospital Arachnoid cysts, old areas of dead tissue from prior silent strokes, benign bone growths, and sinus disease are other common surprises. Most require no treatment, but they occasionally lead to follow-up imaging or referrals, which can cause anxiety for patients who weren’t expecting additional workup.
The lesson here is that “abnormal” on a radiology report does not always mean “dangerous.” Many incidental findings have been quietly sitting in your skull for years and will never cause symptoms. If your report mentions something you don’t recognize, ask your doctor specifically whether it needs follow-up or is simply a normal variant.
When CT Falls Short and MRI Steps In
CT is fast, widely available, and excellent for detecting bleeding, fractures, large masses, and hydrocephalus. But it has real weaknesses. As discussed earlier, early ischemic stroke can be invisible on a plain scan, especially in the posterior fossa. CT also provides less detail about the brain’s soft tissue than MRI does, making it harder to characterize small tumors, demyelinating diseases like multiple sclerosis, infections like encephalitis, or subtle structural abnormalities. For these conditions, MRI with its superior contrast resolution between different soft tissues is the preferred tool.
The trade-off is time and access. An MRI of the brain typically takes 20 to 45 minutes, during which the patient must lie still inside a narrow tube. CT takes under a minute of scanning. In a trauma bay where a patient is unstable, or in a rural hospital where the nearest MRI machine is hours away, CT is the practical choice. The two tests complement each other rather than competing, and knowing what CT can and cannot show helps doctors decide when to stop at CT and when to push for MRI.
Contrast Dye and Safety
Many brain CTs are done without contrast, meaning no injection is needed. But when doctors want to see blood vessels, evaluate for tumors, or look for infection, they inject an iodinated contrast dye through an IV. The dye highlights areas where blood vessels are dense or leaky, making abnormalities stand out against the normal tissue background.
Modern iodinated contrast agents are non-ionic and low-osmolality, and the vast majority of people tolerate them without trouble. When adverse reactions do occur, most are mild symptoms like warmth, nausea, or hives that resolve with observation or a simple antihistamine. However, rare and unpredictable hypersensitivity reactions, including potentially life-threatening anaphylaxis, can still happen.14PubMed Central. Hypersensitivity Reactions to Iodinated Contrast Media People with a history of prior contrast reactions or severe allergies are typically pre-medicated with steroids and antihistamines. Kidney function is also checked beforehand, because the contrast is filtered through the kidneys and can occasionally worsen kidney problems in people who already have impaired function.
Radiation Dose Considerations
Unlike MRI, CT uses ionizing radiation, and the brain sits close to radiation-sensitive structures like the eye lenses. A standard adult head CT delivers a dose in the range of roughly 1 to 2 millisieverts, which is low compared to a chest or abdominal CT but not trivial if scans are repeated frequently. Patients with shunted hydrocephalus, for example, may undergo dozens of head CTs over their lifetime for follow-up evaluations of ventricular size and shunt function.15American Journal of Neuroradiology. Low-Dose Nonenhanced Head CT Protocol for Follow-Up Evaluation of Children with Ventriculoperitoneal Shunt: Reduction of Radiation and Effect on Image Quality
Reducing dose without losing diagnostic quality is an active area of research. One approach is the half-dose method, which lowers the radiation output for follow-up scans. After implementing this in about 28% of head CTs, one hospital achieved roughly a 15% reduction in average radiation dose with no images rated as difficult or impossible to evaluate.16PubMed Central. Evaluation of radiation dose reduction in head CT using the half-dose method For children, the stakes are higher because growing tissues are more sensitive to radiation and kids have more years ahead for any theoretical risk to play out. Studies in pediatric head CT have achieved dramatic dose reductions using modern reconstruction software, with some specialized protocols cutting effective dose by over 95% for specific conditions like craniosynostosis.17PubMed Central. Systematic Review and Meta-Analysis of Radiation Dose Reduction Studies in Pediatric Head CT Newer approaches that tailor radiation output to the individual patient’s head size, rather than using age-based presets, have also shown meaningful reductions, especially in infants and older children.18PubMed Central. Radiation dose optimisation in paediatric head CT using attenuation-based auto prescription
None of this means you should refuse a medically necessary CT scan. The risk from missing a brain bleed or a tumor far outweighs the small radiation exposure from a single scan. The dose conversation becomes relevant mainly when someone needs repeated scans over months or years, which is where low-dose protocols earn their keep.
Artifacts That Can Muddy the Picture
Not every bright or dark spot on a brain CT is a real finding. Artifacts, which are image distortions caused by the physics of the scan or patient factors, can mimic or hide genuine abnormalities. Metal objects like dental fillings, surgical clips, or bullet fragments create streak artifacts that can obscure nearby brain tissue. Motion from a restless or confused patient causes blurring. Beam hardening, a phenomenon where X-rays passing through thick bone get selectively filtered, produces dark bands across the posterior fossa and the base of the skull, and this is partly why that region is so hard to evaluate on CT.
A survey of radiology professionals found that metal artifacts and motion artifacts were correctly identified most often (around 86% and 81% of the time, respectively), while beam hardening artifacts were recognized less reliably, at about 62%. Less experienced professionals had a notably higher rate of incorrect responses.19PubMed Central. Investigation of radiology professionals’ awareness of CT head artifacts This is a useful reminder that reading a CT scan is a skill, and the quality of the interpretation depends on who is reading it. If you’ve ever seen the phrase “limited study due to motion artifact” on a report, it means the images were degraded enough that the radiologist couldn’t be fully confident in their assessment.
Post-Surgical and Device Follow-Up
CT is also the standard tool for checking on hardware inside the head after surgery. Patients with ventriculoperitoneal shunts, metal plates from a cranioplasty, or coils placed inside an aneurysm all need periodic imaging. CT shows the position of the device clearly and can reveal complications like shunt catheter displacement, hardware migration, or new fluid collections around the surgical site. For shunt patients in particular, serial CT scans comparing ventricular size over time are used to decide whether the shunt is working properly or needs revision. The challenge, as noted above, is cumulative radiation exposure when these scans are repeated many times, which has pushed the development of low-dose follow-up protocols.
MRI can sometimes serve as an alternative for follow-up, but many implanted devices are not MRI-compatible, or the metal creates so much signal distortion on MRI that the images are useless around the hardware. CT handles metal much better in this regard, making it the more practical choice despite the radiation trade-off.