I think you should.
If I would want to jump out of topic, yes, but I not want, if other topic is where you extend a question, maybe you should post there?
I probably didn't use the right terminology, but I am talking about image formats which embed information in the coordinates so that a click in at some point results in some embed object responding in a particular manner.
Never heard of, give an example about images that automagical respond to a click.
From my point of view, an image is 2D and static but you can of course have lots of layers on, however, when its rendered it is again just 2D and always static.
Non-Static is in my humble opinion only code that react to something, like a mouse that click inside a specific x/y coordinate.
What I have done in past is to create a base image and alphablended layers (seperate transparent images), when clicked inside a specific coordinate, a layer was added or removed.
(In my case it was a basic landmap with different regions, clicking one would add a visual border around or remove it (alphablended layer on/off)) but all that happen by code and not by image,
but in your case, the way how you describe it, how should an image know anything about a mouse and magicaly react to a click... riddles after riddles.
My suggestion would be to do same as I did, create a base image where you can blend other images in and out.