Temporal Bone CT
- Temporal bone CT is a high-resolution, thin-slice, non-contrast CT built to show tiny bony structures — the ear is basically a fancy machine carved out of bone.
- It runs on the bone algorithm (a sharp reconstruction kernel), reviewed in a bone window, with sub-millimeter slices reformatted in axial and coronal planes.
- It's the workhorse for conductive hearing loss, suspected cholesteatoma, otosclerosis, and bony trauma — anything where you need to count tiny bones and trace tiny canals.
- It is not the test for soft-tissue questions like inner-ear nerve tumors; that's an MRI problem.
- Contrast is usually pointless here — bone doesn't enhance, and the structures of interest are bone, air, and fluid.
The temporal bone is, no exaggeration, the most intricately packed cubic inch in the human body. Crammed inside are three tiny bones that conduct sound, a snail-shaped organ of hearing, a set of looping balance canals, and a nerve that controls your whole face — all threaded through bone with the tolerances of a Swiss watch. To image something this small and this bony, an ordinary CT just won't cut it. You need a CT tuned like a microscope.
What it actually is
Temporal bone CT is a non-contrast CT acquired at very thin slices and reconstructed with a sharp "bone" algorithm. That's the whole secret in one sentence — but each of those words is doing real work.
"Very thin slices" means sub-millimeter (think roughly 0.6 mm or finer). The ossicles — the malleus, incus, and stapes — are smaller than grains of rice. If your slices are too fat, those bones smear together into a useless blur, like trying to read fine print through a steamed-up shower door.
"Sharp bone algorithm" is the reconstruction kernel — the recipe the scanner uses to turn raw data into a picture. A bone kernel deliberately cranks up edge sharpness so the line between bone and air snaps into crisp focus. It adds a bit of grainy noise as the price of that sharpness, but for cortical bone edges, sharpness wins every time. (This same machinery is covered in CT physics and reconstruction.)
Modern scanners acquire one thin-slice volume and then reconstruct it twice — once with a soft kernel for any nearby soft tissue, and once with the sharp bone kernel for the ear. One trip through the scanner, two flavors of image. The patient doesn't get scanned twice.
Why the windowing matters so much
Even a perfect acquisition is useless if you look at it wrong. Temporal bone CT lives in a bone window — a very wide window that lets you tell dense cortical bone apart from slightly-less-dense bone, while air stays pitch black.
Here's the analogy: windowing is like the brightness-and-contrast knobs on an old TV. A "soft tissue" setting spreads all its contrast across the narrow gray range of muscle and fat, which makes bone a featureless white slab. The bone window throws that contrast wide open across the dense end, so you can suddenly see texture inside the bone — the wispy walls of the air cells, the thin shell over a balance canal. Same data, completely different visibility, just by turning the knobs. If you want the deeper version of this, see Hounsfield units and windowing.
A "normal-looking" temporal bone CT viewed on a soft-tissue window is a trap. The pathology — an eroded ossicle, a dehiscent canal, a hairline fracture — often only declares itself in the bone window. Always confirm you're reading bone windows before calling a study clean.
What you're actually hunting for
The strength of this study is bony detail and the air-versus-not-air contrast of the middle ear. The middle ear is normally an air-filled box; fluid, soft tissue, or an eroding mass shows up as something gray filling that black space.
A few classic jobs it does well:
| Clinical question | What the CT shows |
|---|---|
| Conductive hearing loss | Ossicular erosion or fixation; otosclerotic bone changes |
| Suspected cholesteatoma | Soft-tissue mass with bony erosion of ossicles or the scutum |
| Trauma | Temporal bone fractures and ossicular disruption |
| Cochlear implant work-up | Cochlear patency and inner-ear bony anatomy |
Where it runs out of road
CT is brilliant at bone and useless at nuance inside soft tissue. It cannot tell you whether the facial nerve itself is inflamed, or whether there's a small tumor on the hearing-and-balance nerve in the canal. Those are MRI questions — MRI sees the nerve; CT sees only the bony tunnel the nerve runs through.
Rule of thumb: bone, air, and fluid in the bony labyrinth and middle ear are CT problems. Nerves, membranes, and inner-ear soft tissue are MRI problems. The two studies are teammates, not rivals — for the full clinical picture, see Temporal Bone (Hearing Loss, Cholesteatoma).
And because the target is bone — which doesn't enhance — contrast is usually skipped. You only add it when the question shifts toward an enhancing mass or infection spreading beyond bone, and even then MRI is often the better tool.
The one-sentence takeaway
Temporal bone CT is a non-contrast, sub-millimeter, sharp-kernel scan read in a bone window — a microscope for bone — and its single most important habit is this: look at it in the right window, or you'll miss the very thing you scanned for.