CT Protocol
- Multidetector CT ≤1 mm section thickness, sharp (bone) kernel, multiplanar reconstruction
- Always compare with the contralateral side
- MRI useful for intralabyrinthine hemorrhage when CT is negative
Classification
| Feature | Otic Capsule–Sparing | Otic Capsule–Violating |
|---|---|---|
| Frequency | Common — majority of cases | 5–20% |
| Fracture path | Squamous temporal bone, mastoid air cells, tegmen mastoideum/tympani, EAC; spares cochlea/vestibule/SCCs | Through cochlea, vestibule, or SCCs; commonly involves IAC, jugular foramen, facial canal, carotid canal |
| Hearing loss | CHL or mixed (hemotympanum, ossicular injury) | SNHL, vertigo, tinnitus — often profound and permanent |
| Facial nerve palsy | Less common | 4× more likely |
| CSF otorrhea | Less common | 7× more likely |
| Associated injuries | Ossicular chain injury, hemotympanum; EDH if squamous temporal involved | Intralabyrinthine hemorrhage, CN VII palsy, CSF leak; sigmoid sinus / ICA injury with medial extension |
Fracture Line Mimics — Normal Structures
These normal structures should not be mistaken for fractures:
| Structure | Location / Appearance |
|---|---|
| Cochlear aqueduct | Narrow canal caudal and parallel to IAC |
| Vestibular aqueduct | Behind posterior SCC; slitlike |
| Subarcuate canaliculus | Upper otic capsule region |
| Singular canal | Arises posterior to IAC |
| Petro-occipital fissure | Between petrous temporal and occipital bones; originates at jugular foramen |
| Tympanosquamous fissure | Between tympanic and squamous parts |
| Petrosphenoidal fissure | Between sphenoid and petrous temporal bones |
| Cochlear cleft | Anterior to oval window; curvilinear lucency; seen in 41% of pediatric CT |
Key Structure Findings
| Structure | CT/MRI Findings | Clinical Significance |
|---|---|---|
| Tympanic cavity / hemotympanum | Fluid in tympanic cavity and mastoid air cells; high attenuation on CT; high T1 MRI signal = hemorrhage; air-fluid levels | CHL (usually transient with isolated hemotympanum); associated with ossicular injuries; indistinguishable from serous effusion on CT alone |
| Ossicular dislocation | Incudostapedial joint most common dislocation site; incudomalleolar = "ice cream falling off the cone" on axial CT (malleal head = ice cream, incudal body = cone); widened joint space; assessment difficult with surrounding hemotympanum | CHL persisting >2 months warrants dedicated CT; incudomalleolar dislocation occurs with longitudinal Fx |
| Ossicular fracture | Incus long process most vulnerable (fragile, unsupported); stapes crura next; cortical discontinuity on thin-section CT | Persistent CHL >2 months; surgical ossiculoplasty |
| Otic capsule / labyrinth | Fracture through cochlea/vestibule/SCCs; pneumolabyrinth (air within labyrinth on bone kernel CT) = pathognomonic for otic capsule violation even without a visible fracture line; T1 MRI high signal = hemorrhage; FLAIR more sensitive than T1 | SNHL, vertigo, tinnitus; late: labyrinthitis ossificans (faint high attenuation in membranous labyrinth = permanent SNHL); cochlear implant still possible with focal ossificans |
| Facial canal | Transverse Fx → labyrinthine segment + geniculate ganglion most common; longitudinal Fx → geniculate/tympanic segment; look for bony spicules along nerve course | Immediate + complete palsy = suspect transection/compression; look for osseous fragment — surgical decompression target; delayed/incomplete palsy = usually edema, treat conservatively with steroids |
| Tegmen tympani | Best on thin-section coronal CT (1 mm, sharp kernel); pneumocephalus; cephalocele (herniation of brain/dura through defect) | CSF otorrhea; meningitis risk ~10% with CSF leak; surgical repair for persistent leak; MRI best for cephalocele detection |
| Carotid canal | Fracture line interrupting carotid canal walls; air in canal; intramural thrombus = low-attenuation crescentic lesion on CTA | ICA injury in 35% when carotid canal fractured; recommend CTA or MRA; endovascular therapy / anticoagulation as indicated |
| Sigmoid sinus / jugular bulb | Fracture extending to posterior temporal bone; acute thrombus = high attenuation on NECT in sinus lumen; absent flow void on MRI | Venous infarction; contrast CT distinguishes EDH from sinus thrombosis (sinus enhances; EDH does not) |
| EAC / TMJ | Longitudinal Fx → EAC fracture (otorrhagia); tympanic plate fracture → EAC stenosis; TMJ: fracture line and air in joint | Trismus (masseter/TMJ involvement); EAC stenosis; cutaneous entrapment in Fx → acquired cholesteatoma (late complication) |
| Perilymphatic fistula | Oval or round window rupture; CT often negative; may show pneumolabyrinth or unexplained middle ear fluid | Fluctuating SNHL, vertigo worse with Valsalva; most common cause = oval/round window rupture from barotrauma or direct trauma |
Intracranial injuries occur in up to 90% of temporal bone fractures — search for EDH (middle meningeal artery), traumatic SAH, SDH, brain contusion, and pneumocephalus on every case. If the carotid canal is fractured, recommend CTA or MRA to evaluate for ICA injury (present in 35%).
Reporting Checklist — Temporal Bone Trauma
- Fracture type: otic capsule-sparing / otic capsule-violating / bilateral
- Fracture extent: course and specific structures traversed
- Tympanic cavity: hemotympanum present / absent
- Ossicular chain:
- Incudostapedial joint — intact / dislocated
- Incudomalleolar joint — intact / dislocated (ice cream cone sign present / absent)
- Ossicular fracture — yes / no
- Otic capsule: spared / violated; pneumolabyrinth present / absent
- Facial canal: involved / uninvolved; bony spicule along nerve course yes / no
- Tegmen tympani: intact / fractured; pneumocephalus present / absent; cephalocele present / absent
- Carotid canal: intact / fractured — if fractured, RECOMMEND CTA or MRA
- Sigmoid sinus / jugular bulb: involved / uninvolved
- EAC: normal / fractured / hematoma; TMJ: normal / involved
- Intracranial injury: EDH / SAH / contusion / subdural / pneumocephalus / normal
Reference
Kurihara YY et al. Temporal Bone Trauma: Typical CT and MRI Appearances. RadioGraphics. 2020;40(4).