Select a rehabilitation specialty, trigger stream capture, and inspect the clinical structure outputs.
Select your practice discipline from the header bar, then click start capture below to begin dialogue ingestion.
The model tags speaker labels (Therapist/Patient) and timestamps based on audio gaps, creating a structured text feed.
General discussion elements (sports, greetings, stories) are flagged and dimmed, maintaining high documentation density.
Named Entity Recognition targets muscle metrics, joint angles, functional restrictions, and outcomes scales, building an entity rail.