How to predict the generated caption from a new medical image by using the pre-trained model?

[RunCode90]

Hi, I am following your work and thank your for the open code!

But one question is:

If i want to predict the generated caption from a new medical image by using the pre-trained model (mimic_cxr or iu_xray), what should i do?

Is it directly inputting a new image into this pre-trained model?

Can you provide a code example?

Hope you can help.
Looking forward to your reply.
Thank you!
Best wishes!