# Head CT (ct/head)

> Head CT read: bleed volumes, ASPECTS, vessel occlusion, perfusion core and skull fractures.

Upload a non-contrast head CT study as received from the scanner or PACS. The run picks the brain reconstruction, corrects gantry tilt, reads thin series as 5 mm slabs, and answers the first question of every emergency head CT: is there an acute intracranial hemorrhage. Each compartment (epidural, subdural, subarachnoid, intraparenchymal, intraventricular) is stated present or absent; present bleeds come with their outline, side, volume, subdural thickness and the ABC/2 cross-check, and a critical / urgent / routine triage flag leads the result. Outlines return as DICOM SEG on your slices, measurements as a TID 1500 SR.

Model page: https://medrun.ai/models/ct/head
Status: beta · version 1.0.0 · viewer: volume

## Input

A non-contrast head CT study: the DICOM folder, a CD export or a zip of it (all series; the brain series is chosen for you). Thin and 5 mm reconstructions both work.

Accepted: DICOM files or folder, CD export (DICOMDIR), zip.

## Pricing

- `ct/head/interpret`: $0.08 per request
- Draft report (opt-in, `options.report`): +$0.02 per run

## Result sections

- `#triage` Critical-finding triage (available): Worklist priority from the hemorrhage read (critical when a hemorrhage is present, urgent when it cannot be excluded, routine otherwise).
- `#hemorrhage` Intracranial hemorrhage (available): Acute intracranial hemorrhage present, not excluded or absent; each compartment (epidural, subdural, subarachnoid, intraparenchymal, intraventricular) stated; present bleeds outlined in 3D with side, volume, subdural thickness, the ABC/2 cross-check and the ICH score imaging inputs.
- `#midline-shift` Midline shift (planned): Septum pellucidum deviation from the ideal midline in mm, with the > 5 mm flag and direction.
- `#ventricles` Ventricles and hydrocephalus (planned): Ventricle volumes, Evans index and temporal-horn dilatation.
- `#fractures` Skull and facial fractures (planned): Vault, skull-base and facial fractures, depressed fracture and pneumocephalus on the bone reconstruction.
- `#aspects` ASPECTS (planned): Early ischaemic change per ASPECTS region and side in a stroke study without hemorrhage.
- `#lvo` Large-vessel occlusion (planned): Occlusion site and side on a head and neck CTA.
- `#aneurysm` Intracranial aneurysm (planned): Aneurysm detection, location and maximal diameter on CTA.
- `#perfusion` CT perfusion (planned): CBF, CBV, MTT, Tmax maps with core and hypoperfusion volumes and the mismatch ratio.
- `#segment-vessels` Circle of Willis segmentation (planned): Multiclass Circle of Willis segmentation with a centreline graph.
- `#compare` Comparison with prior (planned): Hematoma expansion and new bleeds against a prior head CT.
- `#report` Draft report (available): Opt-in draft report (findings, impression, limitations) written from the structured result; every number is validated against it.

## Endpoint `ct/head/interpret` (run)

The full read of the head CT study: the non-contrast brain series is analysed and the result contains all sections the study supports.

Docs: https://medrun.ai/models/ct/head/api · OpenAPI: https://api.medrun.ai/v1/models/ct/head/interpret/openapi.json

### Input schema

- `files` (array; required): The head CT study: DICOM files, a folder, a CD export or a zip (all series are fine). `file_…` ids from /v1/uploads or https URLs.
- `clinical_context` (object): Optional. Used for guidance only; never names, IDs or dates.
  - `age_years` (integer): Age (years)
  - `sex` (string; one of female, male, other, unknown): Sex
  - `indication` (string): For example "head trauma, on anticoagulation" or "sudden headache".
- `options` (object): Options
  - `report` (boolean; default false): Adds a draft head CT report written from the structured result (English). +$0.02 per run.
  - `return` (array; default ["json","dicom-sr","dicom-seg"]): JSON is always returned. DICOM objects reference your original UIDs.

### Example (cURL)

```bash
response=$(curl --request POST \
  --url https://api.medrun.ai/v1/run/ct/head/interpret \
  --header "Authorization: Bearer $MEDRUN_KEY" \
  --header "Content-Type: application/json" \
  --data '{
  "input": {
    "files": [
      "file_01K7CTHEAD0000000000000000"
    ],
    "options": {
      "report": false
    }
  }
}')
REQUEST_ID=$(echo "$response" | jq -r .id)
```

## Output

The MedRun result envelope: https://medrun.ai/docs/result-envelope