7High-Resolution CT of the Temporal Bone

FSeeing the ear before you open it

The middle ear is a space the surgeon cannot see until the tympanic membrane is lifted, and even then the view is partial: the attic, the sinus tympani and the medial wall hide their contents behind overhanging bone. High-resolution computed tomography (HRCT) of the temporal bone is the one tool that looks through that bone in advance. It cannot tell the surgeon whether the ossicular chain conducts sound, and it cannot grade the mucosa — but it maps the territory: which ossicles are eroded, where the cholesteatoma has eaten into bone, whether the roof and the facial canal are intact, and how well the mastoid breathes [2015].

For ossiculoplasty this matters because the operation is, before anything else, an exercise in anticipation. The choice between a partial and a total prosthesis, the decision to stage a cholesteatoma ear, the warning to slow down over the oval window — each is shaped by what the scan shows. A useful way to hold the right attitude is this: HRCT is a roadmap of the disease and a list of hazards to confirm, not a verdict that replaces the operative finding. The scan tells you where to look and what to fear; the microscope tells you the truth [2020].

FHow the study is acquired and read

A temporal bone HRCT is acquired as a thin-section volume — submillimetre slices, ideally 0.6 mm or less— reconstructed with a high-frequency (bone) algorithm that sharpens the cortical edges at the cost of soft-tissue smoothness. From the volume the radiologist reconstructs the two planes that the otologist thinks in: the axial plane, which lays out the ossicles and the labyrinth from above, and the coronal plane, which shows the scutum, the tegmen, the oval window and the facial canal in the orientation the surgeon meets them [2015]. Multiplanar and oblique reconstructions — along the long process of the incus, or through the stapes — are now routine and turn a stack of slices into a model of the chain.

Reading the study is a discipline of going through the same checklist every time, because the eye that hunts only for the obvious cholesteatoma misses the dehiscent tegmen above it. A workable sequence runs from lateral to medial and from front to back: the external canal and scutum; the ossicular chain joint by joint; the attic, aditus and antrum; the tegmen; the medial wall with its oval and round windows; the facial canal along its tympanic and mastoid course; the labyrinth; and finally the mastoid air cells. Crucially, the reader must know what a normal ear looks like, because some structures are simply not seen in health: in 75 normal middle ears the lateral malleal ligament was visible in 95% but the stapedius tendon in only 27%, so non-visualisation of a fine structure is not evidence of disease [1997].

TReading the ossicles, scutum and tegmen

The ossicular chain is the heart of the ossiculoplasty scan. HRCT assesses the integrity and alignmentof the malleus, incus and stapes; it looks for erosion of the long process of the incus — the commonest single point of failure in chronic otitis media — and for disarticulation at the incudo-stapedial or incudo-malleal joints. A clear gap between two ossicular components on a good reconstruction suggests discontinuity, whereas dense sclerotic material in the oval window niche raises fixation or tympanosclerosis [2020]. But the trainee must learn the scan’s blind spots. The long process and the stapes superstructure are thin, obliquely oriented and small relative to the voxel, so they are precisely the structures CT most often under-reads. In HRCT-versus-operative series the malleus is detected with high sensitivity, the incus less so, and the stapes least of all — which is why a “normal” comment on the incus long process does not certify continuity [2020].

Around the chain, two bony landmarks carry disproportionate weight. The scutum— the sharp spur of bone forming the lateral wall of the attic — blunts and erodes early in attic cholesteatoma, and its loss on a coronal image is often the first hard sign of the disease, guiding the decision between an atticotomy and a wider approach. The tegmen tympani, the thin plate roofing the attic, must be traced for dehiscence: a defect warns of an exposed or herniated dura before the surgeon drills toward it. By contrast, ossicular fixation and tympanosclerosis are read indirectly, from high-density foci at the windows and around the ossicles; CT can flag dense sclerotic change reliably, but it cannot measure whether a footplate actually moves, and reported sensitivities for fixation are high only because the dense focus is conspicuous, not because mobility is being assessed [2021].

TFacial canal, fistula and the limits of CT

Two findings on the medial wall change how an operation is conducted, and both must be hunted for deliberately. The first is dehiscence of the facial canal, most often of the tympanic segment over the oval window, where a curette in disease can meet bare nerve. The second is a labyrinthine fistula— classically a notch in the cortex of the lateral semicircular canal — which mandates extreme care and may change the order of disease clearance. Here CT performs unevenly. For the semicircular-canal fistula it is excellent: in one mastoidectomy series CT and surgery agreed in 59 of 61 ears. For the facial canal it is far weaker — in the same study the sensitivity for canal dehiscence was only about 66%, with a specificity of 84% [1996].

The practical lesson is one of the most important in this whole module: a normal-looking facial canal on CT does not exclude a dehiscent nerve. Thin bone below the voxel limit can look continuous when it is not, and partial-volume averaging can equally over-call dehiscence where bone is merely thin. Imaging therefore flags the high-risk ear — the revision, the cholesteatomatous, the ear with a fistula — but it never licenses a careless curette over the oval window. The operative view always overrides the scan [2020]. Beyond the facial nerve, this is the general weakness of CT: it is a bone study. It cannot distinguish cholesteatoma from granulation or simple fluid in an opacified middle ear, cannot show adhesions or the state of the mucosa, and cannot grade the very soft-tissue factors that most strongly predict ossiculoplasty success. Where that distinction matters — particularly residual or recurrent cholesteatoma behind an intact drum — diffusion-weighted MRI, not more CT, is the answer [2015].

TAeration as a prognostic sign

One of the most useful pieces of information the scan offers is also the easiest to overlook: how well the ear is aerated. The middle ear is an acoustic transformer that works only when it is an air-filled space, and the mastoid air-cell system is both a reservoir of that air and a record of the ear’s ventilation history. On HRCT the reader notes the degree of pneumatisation, the presence of sclerosis, and whether the middle-ear cleft and antrum are clear or opacified. A well-pneumatised, clear system signals an ear that has ventilated well; a contracted, sclerotic, opacified system signals the opposite [2020].

This is not merely descriptive, because the state of the middle-ear environment is among the strongest determinants of how an ossiculoplasty will fare. In the OOPS staging system derived from 200 ossiculoplasties, mucosal and middle-ear status were independent predictors of the hearing result, outweighing many technical variables — a finding echoed across the literature that a hostile, poorly ventilated ear undermines even a technically perfect reconstruction [2001]. A sclerotic, opacified mastoid on the preoperative scan therefore does two things: it tempers the prognosis the surgeon offers, and it strengthens the case for staging— clearing disease and restoring a dry, ventilated ear first, and reconstructing the chain at a second sitting once the environment has improved.

CFrom scan to surgical plan

Read as a whole, the temporal bone HRCT answers a short series of operational questions, and the clinician’s skill lies in turning each answer into a decision rather than a description. Is there disease, and how far has it spread? The extent of soft tissue and bony erosion sets the approach and the likely need to stage. What is the ossicular chain likely to need?An eroded long process points toward a partial prosthesis onto the stapes head, an absent superstructure toward a total prosthesis onto the footplate — provisional plans, to be confirmed by hand at operation [2020]. Where are the hazards?A tegmen defect, a possible facial canal dehiscence, a lateral canal fistula — each is a place to slow down, and the facial canal in particular must be treated as potentially bare whatever the scan says [1996]. How favourable is the environment? Aeration and sclerosis calibrate the prognosis and the decision to stage [2001].

In the altered anatomy of revision and canal-wall-down ears, the scan earns its keep again, mapping the obliterated mesotympanum, the lowered facial ridge and the residual cells before a second dissection. And throughout, three disciplines keep the imaging honest. CT findings are correlated with otoscopy and audiometry, never read in isolation: a scan that disagrees with a 40 dB conductive gap behind an intact drum should prompt a second look, not blind faith. MRI is added, not substituted, when soft tissue must be characterised — diffusion-weighted sequences for suspected residual cholesteatoma in particular [2015]. And the operative finding remains the gold standard for ossicular continuity, fixation and the bareness of the facial nerve, because those are judgements of touch and of light that no current scanner can make[2021]. The surgeon who holds the scan in one hand and this scepticism in the other walks into the ear already knowing the country — and ready to be corrected by it.