Please enable JavaScript.

Coggle requires JavaScript to display documents.

GenRx Functionality, Unresolved Threads - Coggle Diagram

- - - - Student enters answers in a text entry box or uses the mic and speech to text
        
        Student submissions are sent via API to OpenAI, responses are received and displayed
        
        Students can use text to speech to listen to responses or just read
    - - I'm thinking genAI image over trying to do some live animated lipsync video - that would be alienating I think
  - - - Click hotspot, see information, log completion of activity (optional)
  - - - On submission show model answer
        
        Optional - could be some highlighting or analysis of the student's submitted answer?
    - - See question, choose a response, show correct/incorrect
    - - Maybe if we make the patient say something that conflicts with the documentation there's an option to update documentation somehow - maybe this is just a text field
        
        All these responses are collected ideally for the final debrief. If possible then submitted to Canvas
    - - See a selection of images and pick one, like an MCQ just with media
- - - - cases act as a sub site
        
        Cases can contain pages that each contain content, like VIC
        
        Content types
        
        Patient Interaction
        
        Patient image
        
        Simple image upload
        
        1 more item...
        
        Persona Management
        
        This is where you'd update the prompt, text field as simplest approach for OpenAI API
        
        Potentially here you'd add 'sources of truth' too to the patient like charts or other assets
        
        1 more item...
        
        Patient Personas created with SPIIF
        
        2 more items...
        
        Physical Assessment
        
        Hotspots on video, images etc
        
        Easy start option is H5P, otherwise it's custom JS
        
        1 more item...
        
        DTP/Care Plan Flow
        
        Text questions with model answer
        
        MCQs
        
        Identify discrepencies/update chart
        
        This would be heavily modelled off of the current VIC DTP and Care Plan flow but with more flexibility ideally, due to the CMS
        
        1 more item...
        
        Image selection
        
        Documentation pages (like VIC)
        
        Text, media, click to view
        
        Text fields to add information, media upload or WYSIWYG fields to create information to display on click
        
        1 more item...
        
        The url is the same but each case has got its own endpoint like ubcpharmvic.com/case-a, ubcpharmvic.com/case-b
        
        users should be granted access to sub sites individually
      - Deleted cases are saved in trash for 90 days or longer
        
        Periodic backups?
        
        JP indicates this is dictated by the privacy office
      - If LTI enabled no deletion available if a case is associated with Canvas
      - Case-wide settings include whether to track time and money cost (this controls whether those fields show)
  - - - PDF, jpg, mp4, etc
        
        Re use materials on different pages
        
        what about embeds like youtube videos or anatomy.tv
  - - - Different roles? Editor vs admin?