Listen
This endpoint is used to transcribe audio files and/or generate clinical notes from them. The API call is a POST request to
v1/scribe/user/listen or v1/scribe/server/listen
You can check out the API documentation at Listen.
Input
The Listend API call requires a multi-part form data input. The first input is a binary audio file.
The second input is a listen_request object which consists of the following parameters:
| Parameter | Values | Description |
|---|---|---|
| output_objects (required) | Output Objects (string) or Array of Output Objects (any) (Output Objects) | Output to be returned from the API (one or both of transcript/note) |
| one of: "transcript", "note", ["transcript", "note"] | ||
| language (required) | string (Language) Enum: "en-US" "fr-CA" | Language of consultation |
| note_template (required) | string (Note Template) Enum: "soap" "consult" | Clinical note template to use (e.g. soap, consult) |
| style (required) | string (Style) Enum: "paragraph" "bullet" "auto" | Formatting style to use for each section in the clinical note |
| note_gen_mode (required) | string (Note Gen Mode) Enum: "fast" "medium" "best" | Quality of note generation |
| transcribe_gen_mode (required) | string (Transcribe Gen Mode) Enum: "fast" "medium" "best" | Quality of transcription |
Output
The API returns a ScribeResponse JSON object which consists of the following parameters: transcript and/or note or both. See scribe-outputs for more details.