Have you tried using ABBYY Finereader to "read" the pdf?
PDF is not really a good source for conversion. A lot of mess comes up with PDF as a source. Split text at the bottom of the page, missing letters, etc.
My solution to that is to process PDF through OCR software like ABBYY. Then save it in DOCX or RTF. It even retains the formatting if you choose Exact Copy option. It will also copy the image but you have to tweak it a little for it to read the WHOLE picture.
