De-identification: software and test data

The MIMIC-II project has contributed software for deidentification of free text in medical records, and a corpus of deidentified text that can be used for development and evaluation of deidentifcation software.