3a: Image-Recognition Plugin (transforms visual information into text for chatbot)

0—completely irrelevant description

1—identifies context of medical topic

2—identifies context of surgical field

3—correctly identifies anatomy or surgical instrument

4—correctly identifies relevant pathology

5—correctly identifies relevant surgical procedure

3b: Chatbot (the one the user interacts with)

0—Apologizes for poor extraction results from image information

1—Chatbot poorly connects poor results to a concrete answer

2—Chatbot correctly connects poor results to a concrete answer

3—Chatbot correctly analyzes relevant anatomy or instrument

4—Chatbot correctly analyzes relevant pathology context

5—Chatbot correctly analyzes relevant surgical procedure