How to read a brain CT scan

What you are actually looking at

A head CT is a stack of axial slices — horizontal cross-sections, taken from the base of the skull upward. By convention they are displayed as though you were standing at the patient’s feet looking up, which means the left side of the brain appears on the right of the image. This trips up almost every first-time reader, and it is the reason a report can say “left frontal” about something that sits on the right of the picture you are looking at.

Each pixel carries a density value in Hounsfield units, on a scale fixed by definition: water is 0, air is about −1000, dense bone is up to around +1000. Fresh clotted blood sits high, in the 50 to 70 range, which is why acute bleeding looks conspicuously bright. Swollen or infarcted tissue holds more water and drops lower, which is why it looks dark. Almost all of the language in a head CT report — hyperdense, hypodense, isodense — is describing where something sits on that scale relative to normal brain.

Why the same scan is looked at three times

The full density range is far wider than a screen or an eye can show at once, so the radiologist views the same data through several windows — a narrow slice of the scale, stretched across the full range of grey. A brain window compresses the narrow band where grey and white matter live so the difference between them becomes visible. A bone window stretches the top of the scale, making fractures obvious while the brain turns into a featureless blur. A subdural or blood window sits in between, so a thin layer of blood lying against the inner skull does not disappear into the brightness of the bone next to it.

This is why a scan can look completely different in two images from the same study, and why viewing a head CT in a single default window misses findings the radiologist saw. It is one setting, not one picture.

The order a radiologist works through

A systematic search pattern is what stops an obvious finding from hiding a subtler one. The specifics vary by training, but the shape is consistent.

1. 1.Confirm orientation and adequacy

   Right side, correct patient, slices covering from the skull base to the vertex, no severe motion. A study that does not cover the whole brain cannot be reported as if it did.

2. 2.Compare the two halves

   The brain is close to symmetric. Reading left against right, slice by slice, is how a subtle difference in brightness or in the size of a space becomes visible, when neither side alone would look wrong.

3. 3.Check the midline

   The structures down the centre should sit down the centre. Any sideways displacement means something on one side is taking up space it should not be. This is the single most consequential observation on an emergency head CT.

4. 4.Look at the grey-white boundary

   Normal cortex is slightly denser than the white matter beneath it, giving a visible border. Loss of that distinction over a region is one of the earliest signs of an evolving infarct or of swelling.

5. 5.Assess the ventricles

   The fluid-filled spaces should be symmetric and appropriately sized for the person’s age. Enlarged can mean fluid is not draining, or simply that surrounding tissue has thinned with age. Compressed or effaced means something is pressing on them.

6. 6.Check the basal cisterns

   The CSF spaces around the brainstem should be open and dark. Effaced cisterns, or bright material filling them, are findings that change management immediately.

7. 7.Sweep the surfaces and the skull

   Along the inner table for thin collections of blood, then the bone window for fractures, and finally the sinuses, mastoids and orbits, which are inside the field of view and get commented on.

Normal versus abnormal, at a glance

Two of those rows are worth separating from the rest. “Age-appropriate” is doing real work: ventricles and sulci widen gradually across a normal lifetime, so a scan that would be unusual at thirty is unremarkable at eighty, and reports describe this as involutional or age-related change rather than as damage. And symmetry has limits — small asymmetries of vessels, ventricles and skull thickness are common in healthy people, which is precisely why an abnormal-looking pixel is not a finding until someone trained decides it is.

What a head CT can miss

• An early ischaemic stroke. In the first hours a CT is frequently normal; it is performed at that stage mainly to exclude bleeding, because that changes treatment immediately. MRI, specifically diffusion-weighted imaging, shows the infarct far sooner.
• Small lesions, and anything near the skull base, where beam hardening from dense bone streaks across the posterior fossa and degrades the image.
• Most of what MRI is for: subtle white-matter disease, early inflammation, small tumours, and marrow.
• Anything outside the scanned volume. A head CT is not a scan of the neck vessels or the cervical spine unless those were requested.

This is why “no acute intracranial abnormality” is such a precisely worded phrase. It says that nothing needing urgent action is visible on this study, on this modality, at this moment. It does not say that nothing is wrong, and it is not a substitute for the clinical assessment that continues after the scan.

What this will and will not let you do

Knowing the checklist above will let you follow your report, ask sharper questions, and stop misreading orientation and brightness. It will not let you read your own scan, and the gap between those two things is larger than it looks. A radiologist has spent years learning which asymmetries are normal variants, how age changes the baseline, and how each finding shifts in meaning depending on the clinical context — and even then, borderline cases are read differently by different experts, which is why formal second opinions exist.

Use this to understand the document you were given. Take the interpretation from the person who ordered it.

Frequently asked questions

Why is the left side of the brain on the right of the image?

Axial CT and MRI slices are displayed as though the viewer were standing at the patient’s feet looking up, so the patient’s left appears on the viewer’s right. Reports always describe left and right from the patient’s point of view, never the image’s, which is why a “left-sided” finding can sit on the right of the picture.

What does bright white mean on a brain CT?

High density. In the brain that usually means fresh clotted blood, calcification, contrast material, or bone. Fresh bleeding is conspicuously bright and is the finding a head CT is best at detecting, which is why CT rather than MRI is the first scan after a head injury or a sudden severe headache.

What does a dark area on a brain CT mean?

Lower density than normal brain, which usually means more water — swelling, an established infarct, an old area of damage, a fluid-filled space, or a cyst. Age alone does not make a dark patch harmless: what it means depends on its shape, location, borders and how it fits your symptoms.

Can a brain CT be normal if something is wrong?

Yes. An ischaemic stroke can leave a CT looking entirely normal for hours, and CT is poor at small lesions, early inflammation and anything close to the skull base. A normal head CT excludes the emergencies CT is good at excluding; it does not exclude everything, and further imaging is often the next step.

What is midline shift and why does it matter?

The structures running down the centre of the brain should sit in the centre. If they are pushed sideways, something on one side is occupying space it should not — bleeding, swelling or a mass. It is one of the most urgent observations on a head CT because it means pressure is being transmitted across the brain.