Feature request: support for "title" and "context" params for ImageBlockParam as well

[SilverLaius]

First off, love to see the new context and title keys for documents, good job!

My use case is data extraction from multiple pdfs and images in a single request. Since we mix images and pdfs together in the same request, it would be useful to also allow adding context and titles to images as well, instead of providing that context in the prompt, mixed up with instructions.