A head CT without contrast produces detailed cross-sectional images of the brain, skull, and surrounding structures by measuring how different tissues absorb X-rays. It excels at detecting acute bleeding inside the skull, skull fractures, shifts in brain position, enlarged fluid-filled spaces, and calcifications. Because it requires no intravenous dye and takes only seconds, it is the standard first-line imaging test in emergency rooms for head injuries, suspected strokes, and sudden severe headaches. Its strengths and blind spots, though, are worth understanding in detail.
How the Scan Creates an Image
A CT scanner rotates an X-ray beam around your head and measures how much each tiny block of tissue absorbs those X-rays. Dense materials like bone absorb a lot and appear bright white. Air absorbs almost nothing and appears black. Brain tissue, blood, fat, and cerebrospinal fluid fall somewhere in between, each at a slightly different shade of gray. These density values are measured on a standardized scale. Normal brain gray matter, for instance, measures in the mid-30s on that scale, while the fluid inside the ventricles measures close to 10, and fresh blood comes in around 50 to 70. That reliable difference in density is what makes the scan so useful for distinguishing one tissue type from another without any injected contrast dye.
Radiologists can adjust the display settings on the same scan data to highlight different structures. A “brain window” setting emphasizes the subtle gray-tone differences between brain tissues. A “bone window” setting cranks up the contrast to make the skull’s fine details visible, which is how fractures and bony lesions are evaluated. No additional scanning is needed for these different views; they come from the same raw data.
Acute Bleeding Is Where Non-Contrast CT Shines
Fresh blood is denser than normal brain tissue, so it lights up as a bright white area on the scan. This makes non-contrast CT extremely sensitive for detecting acute intracranial hemorrhage, which is why it is the go-to test when someone arrives at the emergency department with a sudden neurological change. The scan can distinguish different types of bleeding based on where the bright signal appears: within the brain tissue itself, between the brain and its membranes, or in the fluid spaces.
Non-contrast CT is considered a first-line diagnostic test for acute stroke largely because of its low cost, wide availability, and rapid detection of intracranial hemorrhage.1Seminars in Ultrasound, CT and MRI. Noncontrast CT in Acute Stroke This speed matters enormously because the treatment for a bleeding stroke is the exact opposite of the treatment for a clot-caused stroke. Giving a clot-busting drug to someone who is actually bleeding can be fatal, so ruling out hemorrhage within minutes of arrival is the scan’s most critical job.
Beyond simply detecting a bleed, the scan can sometimes suggest the cause. The location, shape, and surrounding features of the hemorrhage on a non-contrast image can help predict whether the bleeding is due to high blood pressure, a ruptured blood vessel malformation, or something else. One study found that when non-contrast CT images showed features strongly suggesting a non-vascular cause, the negative predictive value for an underlying vascular abnormality was about 95%.2Scientific Reports. The accuracy of non-contrast brain CT scan in predicting the presence of a vascular etiology in patients with primary intracranial hemorrhage Deep brain location combined with a known history of hypertension further improved the ability to predict a hypertensive cause without needing additional vascular imaging.3Journal of Clinical Neuroscience. When less is more: Non-contrast head CT alone to work-up hypertensive intracerebral hemorrhage In practice, many patients with hemorrhage will still get follow-up vascular studies, but the non-contrast scan provides enough initial information to guide urgent decisions.
What the Scan Shows in Ischemic Stroke
When a stroke is caused by a blocked blood vessel rather than a bleed, the scan’s role shifts. In the first few hours, the affected brain tissue may look completely normal, which is actually useful information: a normal-appearing scan in someone with stroke symptoms confirms the problem is a blockage, not a hemorrhage, and clears the way for clot-dissolving treatment. As time passes, though, the scan begins to reveal subtle changes in the oxygen-starved tissue.
These early ischemic signs include a few recognizable patterns. One is the “hyperdense artery” sign, where a clot inside a major brain artery appears as an unusually bright segment of the vessel. Others include blurring of the normal boundary between gray and white matter, subtle darkening of the deep brain structures, and flattening of the brain’s surface grooves.4PubMed Central. Detection of Early Ischemic Changes in Noncontrast CT Head Improved with “Stroke Windows” These changes emerge on a rough timeline: the hyperdense artery sign tends to fade after the first six hours, while tissue darkening becomes more prominent as time goes on. Studies quantifying the extent of this darkening have found it can help estimate how long ago the stroke began, which has practical value when a patient cannot report when their symptoms started.5European Journal of Radiology. Associations between early ischemic signs on non-contrast CT and time since acute ischemic stroke onset: A scoping review
The challenge is that these early signs are genuinely subtle. Even experienced radiologists can miss them, especially in the first few hours. Specialized display settings called “stroke windows” can improve detection by adjusting the contrast levels to make small density differences more apparent. Still, non-contrast CT is far less sensitive to early ischemic damage than MRI, and in many stroke centers, CT angiography (which does use contrast) or CT perfusion is added to the workup to map the blocked vessel and identify salvageable tissue.
Midline Shift and Mass Effect
One of the most clinically urgent findings on a non-contrast head CT is midline shift, where the brain’s central structures are pushed to one side by a mass, swelling, or large blood collection. This displacement is a sign of dangerously increased pressure inside the skull and often triggers emergency surgical intervention.6PubMed Central. Brain Midline Shift Measurement and Its Automation: A Review of Techniques and Algorithms The amount of shift is measured in millimeters on the scan and correlates with the severity of the underlying problem, whether that is a large hemorrhage, a swollen area of dead tissue after a stroke, a tumor, or an abscess.
Measuring midline shift has traditionally been done by hand, but automated tools using artificial intelligence are now reaching accuracy levels comparable to expert neuroradiologists. A multicenter validation study found that one such tool achieved an average measurement error of just 0.8 millimeters, which was actually slightly better than the average difference between pairs of expert readers measuring the same scans.7PubMed. Automated Midline Shift Quantification on Noncontrast CT Across Intracranial Pathologies: A Multicenter Validation Study This kind of automated flagging can speed up triage by alerting clinicians to dangerous shifts before a radiologist has formally read the scan.8PubMed. Efficient automated quantification of midline shift in intracerebral hemorrhage using a binarized deep learning model on non-contrast head CT
Hydrocephalus and Fluid-Space Changes
The brain’s ventricles are fluid-filled chambers that appear dark on CT because cerebrospinal fluid is much less dense than brain tissue. When these ventricles become abnormally enlarged, the scan picks it up readily. The causes range from blockage of fluid flow (obstructive hydrocephalus) to insufficient fluid absorption or brain atrophy that passively expands the spaces.
Distinguishing between these causes matters, and non-contrast CT provides several clues. In normal pressure hydrocephalus, a treatable condition that causes gait problems, urinary incontinence, and cognitive decline in older adults, the imaging features include ventricles that are disproportionately large compared to the brain’s surface grooves, a specific ratio of ventricular width to skull width greater than 0.3, thinning and upward bowing of the structure connecting the brain’s two halves, and widening of particular fissures while the grooves at the top of the brain appear compressed.9PubMed Central. Neuroimaging in normal pressure hydrocephalus Getting this distinction right is important because normal pressure hydrocephalus can be treated with a shunt, while ventricular enlargement from brain atrophy cannot.
Obstructive hydrocephalus, where something physically blocks the flow of cerebrospinal fluid, can be a neurosurgical emergency. Early signs on non-contrast CT include the ventricles expanding unevenly, the third ventricle bulging outward, and fluid seeping through the ventricle walls into surrounding tissue. These findings can be subtle, especially when complicated by a simultaneous hemorrhage or mass. AI-driven triage tools are now being developed to automatically flag suspected obstructive hydrocephalus on non-contrast scans to help clinicians catch it faster.10Stroke. Abstract WP301: Diagnostic Accuracy of an AI-Driven Triage Tool for Obstructive Hydrocephalus on Non-Contrast Head CT
Skull Fractures and Bone Abnormalities
By switching the display to bone window settings, the same scan data reveals the skull in fine detail. Depressed fractures, where a piece of bone is pushed inward, and comminuted fractures, where bone is shattered into pieces, are reliably detected. Fractures at the base of the skull, which are notoriously hard to spot on plain X-rays, are better shown on CT, especially with bone window settings.11Clinical Radiology. Can computed tomography be relied upon to detect skull fractures? The same bone window approach allows identification of skull lesions and bony abnormalities that would be invisible on the standard brain window.12PubMed Central. Window classification of brain CT images in biomedical articles
There is a notable limitation, though. Simple linear fractures that run horizontally, parallel to the plane of the CT slices, can be missed entirely because the fracture line falls between or along the scan slices rather than crossing them. The same study that found depressed and comminuted fractures were all detected reported that none of the horizontal linear fractures were seen, and only about 38% of oblique linear fractures were caught on bone window settings.11Clinical Radiology. Can computed tomography be relied upon to detect skull fractures? Modern scanners with thinner slices have improved this, but horizontal fractures remain the hardest to detect.
Calcifications, Air, and Foreign Objects
Non-contrast CT is uniquely good at showing calcifications because calcium is very dense and appears bright on the scan. The brain normally accumulates calcium deposits in predictable locations as people age, particularly in the pineal gland and a structure called the choroid plexus inside the ventricles. One study of head injury patients found incidental intracranial calcifications in roughly 62% of scans, with the vast majority being benign and expected.13PubMed Central. Incidental cranial CT findings in head injury patients in a Nigerian tertiary hospital Knowing what normal calcification looks like is valuable because abnormal calcification in unexpected locations can signal tumors, infections, vascular malformations, or metabolic disorders.14Radiography Open. Evaluation of intracranial physiological calcifications in Computed Tomography
Air inside the skull, called pneumocephalus, also shows up clearly because air is the least dense substance the scanner encounters and appears jet black. Small amounts of intracranial air after surgery or trauma are common and usually harmless, but larger or increasing collections can indicate a skull base fracture or an ongoing communication between the sinuses and the brain cavity. Foreign objects, particularly metallic ones, are obvious on CT because metals are extremely dense. A case report described how a non-contrast CT clearly identified a fragment of a dagger that had penetrated through the eye socket into the brain cavity, along with the resulting pneumocephalus.15PubMed Central. Occult Orbital Injury with Dagger Fragment with Resulting Pneumocephalus
What Non-Contrast CT Struggles With
For all its strengths, a non-contrast head CT has real blind spots. The most significant is its limited ability to detect brain tumors. Without contrast dye, many tumors blend in with surrounding brain tissue because their density is too similar. A classic study found that in most cases of suspected brain tumor, a single contrast-enhanced scan provided the same diagnostic accuracy as doing both non-contrast and contrast scans, implying that the non-contrast portion alone adds little.16PubMed. Non-contrast CT scanning: limited value in suspected brain tumor Large tumors that cause visible swelling or calcification can still be spotted, but smaller or low-grade tumors may be invisible.
Diffuse axonal injury, a common consequence of traumatic brain injury where nerve fibers are sheared by rotational forces, is another major blind spot. CT often appears normal even when significant axonal damage has occurred. MRI, with its superior ability to distinguish soft tissue contrasts, is far more sensitive for detecting these injuries along with tiny microbleeds and subtle swelling within the brain tissue.17PubMed Central. Comparative Efficacy of MRI and CT in Traumatic Brain Injury: A Systematic Review This is why patients with persistent symptoms after head trauma sometimes have a “normal” CT but an abnormal MRI: the CT was never designed to see that kind of injury.
Brain infections, including abscesses and encephalitis, also fall into the category of conditions where contrast makes a major difference. An abscess may appear as a subtle area of low density on a non-contrast scan, but its characteristic ring enhancement, the bright ring of inflamed tissue that surrounds the pus collection, only becomes visible with contrast. In the early stages of infection, the non-contrast scan may look entirely normal.
Artifacts That Can Mimic Real Disease
CT images are not perfect representations of anatomy. Various artifacts, essentially image distortions caused by the physics of the scanning process or the patient’s anatomy, can create false findings. Dense materials like metal implants, dental fillings, or thick bone at the skull base produce streak artifacts: bright and dark lines that radiate across the image and can obscure or mimic brain lesions.
One reported case described a CT artifact in the back of the brain that closely mimicked a hemorrhage. Without further investigation, such an artifact could have led to unnecessary and potentially harmful treatment.18Journal of Trauma and Injury. Unusual Brain Computed Tomography Artifact in Cerebellum Mimicking Hemorrhage: A Case Report Equipment-related artifacts can also occur. A detector malfunction in a multi-detector scanner can create repeating patterns across images that may be mistaken for actual pathology, prompting recommendations for routine detector calibration.19PubMed. Multi-detector row CT artifacts that mimic disease The posterior fossa, the area near the base of the skull that houses the cerebellum and brainstem, is particularly prone to artifacts because it is surrounded by thick bone. Radiologists are trained to recognize these patterns, but they remain a real source of diagnostic error, especially in emergency settings where speed is prioritized.
Incidental Findings and What to Do About Them
When a head CT is done for one reason, it frequently reveals unrelated findings. In a study of trauma patients, incidental findings appeared on about 85% of scans, with over 90% of those findings being benign.13PubMed Central. Incidental cranial CT findings in head injury patients in a Nigerian tertiary hospital The most common incidental finding was intracranial calcification. Other benign incidentals include small cysts, normal anatomic variants, and sinus disease. Occasionally, the scan reveals something clinically significant that the patient and their doctors did not know about, such as an unsuspected mass or vascular abnormality.
These incidental discoveries create a tricky situation. A benign finding noted in the radiology report can provoke unnecessary anxiety and lead to follow-up imaging, additional costs, and sometimes even invasive procedures that carry their own risks. On the other hand, catching an early-stage significant lesion by chance can be genuinely lifesaving. There is no clean rule for how to handle every incidental finding, but the general principle is that findings with no known clinical consequence should be noted and left alone, while those with potential to cause harm deserve follow-up.
Pediatric Skulls Present Special Challenges
In children, the skull is a work in progress. It is made up of separate bones joined by sutures that gradually fuse over years, and the appearance of these sutures changes with age. On a non-contrast CT, an unfused or partially fused suture can look alarmingly similar to a fracture, and vice versa. Accessory sutures, which are normal developmental variants not present in all children, add another layer of confusion.20PubMed. CT of Normal Developmental and Variant Anatomy of the Pediatric Skull: Distinguishing Trauma from Normality
The stakes run in both directions. A normal suture misread as a fracture can lead to unnecessary hospitalization, child abuse investigations, and parental distress. A real fracture dismissed as a normal suture can mean missed treatment for a serious injury. Familiarity with how the pediatric skull looks at different ages is essential, and the increased resolution of modern scanners helps, but this remains one of the more error-prone areas in emergency radiology.
When the Scan Is Ordered Unnecessarily
Non-contrast head CT is so quick and accessible that it gets ordered frequently for complaints where it rarely finds anything actionable. A large retrospective study in an Italian emergency department classified about a quarter of head CT requests as “usually not appropriate.” Among those inappropriate scans, only 1% showed acute brain disease, compared with 11% among appropriately ordered scans.21PubMed Central. Evaluating head CT referral quality and appropriateness in an Italian emergency department: a monocentric retrospective study The most common reasons for these low-yield scans were headache and fainting, symptoms that by themselves, without additional red-flag features, rarely indicate a condition that non-contrast CT would detect.
Every head CT delivers a dose of ionizing radiation to the brain and eyes. Modern techniques are bringing that dose down. Deep learning-based image reconstruction methods have enabled radiation dose reductions of up to 44% while maintaining diagnostic image quality.22PubMed Central. Deep learning and iterative image reconstruction for head CT: Impact on image quality and radiation dose reduction-Comparative study Still, the cumulative effect of repeated scans over a lifetime, particularly for children and young adults, is a legitimate concern. The best scan is one that is both clinically needed and performed at the lowest dose that still answers the clinical question. If your doctor orders a non-contrast head CT in an emergency, the benefit almost certainly outweighs the radiation risk. If you are being sent for a scan for a chronic, non-urgent headache with no other worrisome features, it is reasonable to ask whether imaging is likely to change your care.
Post-Surgical Monitoring
After brain surgery, non-contrast head CT is the workhorse follow-up imaging tool. It can confirm that a blood collection has been evacuated, show the position of surgical hardware like shunt tubing or drainage catheters, and detect common postoperative complications such as new bleeding, air collections, or worsening swelling. Surgeons typically order serial non-contrast scans in the hours and days after an operation to track changes over time. Expected postoperative findings include small amounts of air within the surgical cavity, minor hemorrhage along the operative tract, and mild local swelling. Distinguishing these routine findings from genuine complications requires comparing each new scan with the previous one, which is another reason the non-contrast approach is favored: the same scan protocol each time makes comparisons straightforward.