Systematic flow for doing word retrieval task

[wqyeo]

Implement the system flow for doing word retrieval task

Main Scenario (Assume correct)

1. Client initiates word retrieval task
2. API sends initial information (picture)
3. Client awaits input from Patient for possible answer
4. Client sends input to API
5. API validates input to be correct
6. Client celebrates for patient

Alternative Scenarios

Patient Send wrong Input

1. Client awaits input from Patient for possible answer
2. Client sends input to API
3. API validates input to be incorrect
4. API sends hint to client. (Hint should be based on the number of hints shown already).
5. Client pops hint

Patient Idles too long; Patient accepts hint

1. Client awaits input from Patient for possible answer
2. Patient idles for a long period of time (~1 minute?)
3. Client asks if Patient needs hint.
4. Patient accepts hint
5. Client requests hint from API
6. API sends hint to client. (Hint should be based on the number of hints shown already).
7. Client pops hint

Considerations

• Save states; Patient might do a task halfway and quit the application.
• Insufficient hints set by practitioner, consider Generating using LLM. See Generate hints for word retrieval practice task #21

---

Related Objectives

• Check correctness of patient's answer to Word Retrevial task Question #28
• API endpoint for transcribing speech #29