As we all know, medical records are particularly challenging for translators; first and foremost because the majority of these documents are handwritten (SOAP notes, progress reports, prescriptions, etc.). Furthermore, many of these handwritten notes are shorthand transcriptions of information dictated by the physician while in the midst of conducting an examination, performing a surgical procedure, examining films, conducting an autopsy, etc.
The rise of e-medical records and, more recently, integration with Tablet PC handwriting recognition systems seemed to be the magic solution for converting obscure handwritten notes into intelligible and grammatically correct text. However, it was recently pointed out to me by a physician that most software fails in that there is no integrated symbol-recognition library. The vast, multilingual character banks do not account for non-character symbols, which are heavily used in medical notes.
Food for thought and certainly a challenge for developers who will need to work to combine mixed text and symbol recognition for each supported language.
Looks like we will all just need to keep clicking our heels as we wait for a solution to the infamous physician handwriting problem.