using experimental (and currently free) software from http://www.asel.udel.edu/speech/DownloadInstr.html you can create a text-to-speech engine using your own voice. However, it looks like it takes a loooong time and many sentences to train the engine correctly.