DIRECTION 02 / DOCTORS
What if AI could work with the context a doctor sees?
A conversation is one signal. The patient’s history, observations, and changing patterns add the rest.
An idea to investigateThe possibility
I’m exploring intelligence over a richer clinical data layer: speech, reports, longitudinal history, and potentially visual observations. The aim is to investigate useful patterns and questions, with clinical judgment staying with the doctor.
The context underneath
What the patient says, what a clinician observes, what the record shows, and what changes over time.
What could be useful?
- Bring relevant clinical context into one view
- Explore patterns across voice, vision, reports, and history
- Surface questions and supporting evidence for a clinician to assess
What I want to learn
- Which observations can be captured reliably and with consent?
- When would a suggestion improve a doctor’s reasoning?
- How should uncertainty, missing evidence, and errors be made visible?
A useful suggestion should be traceable, open to correction, and under clinical control.
These are exploratory directions. The first implementation, its form factor, and any clinical or operational value would need to be tested in the setting where it is used.