OCR is probably the only way. From the PDF perspective, there's a rectangle and some of those rectangles have two lines drawn through them. They're not even images but actual vector drawing commands. You could possibly look for that extra drawing of an "x" but it is unrelated to the text that appears beside it so'd have to write some fuzzy logic to estimate what "x" goes to what "text" and I think you'd end up with a bunch of false positives. If you've got a bunch of these PDFs it might be worth writing something, otherwise OCR or manual entry.
If you want to parse the PDF you can try something like this which is a little ugly but if you're parsing the same PDF over and over again it might work OK. If you want something more generic and reusable I would check out the creator of iText's post here. His post is for optional content groups but it should give you some ideas to start with.