Difference between revisions of "De-identify and anonymize"
| Line 1: | Line 1: | ||
| − | :::'''Return to Main Page''' [[ | + | :::'''Return to Main Page''' [[Main Page]] |
'''De-identification''' removes Private Health Information (PHI) from DICOM files by stripping certain of the DICOM 'tags' from the image headers. There are 1880 DICOM tags but only certain ones need be removed and this can be performed automatically by publicly available software. What remains should include a uniquely assigned study number that allows the image to be linked to other clinical and image data from that same individual. Only the originating institutional investigator holds a code key that could under specially permitted circumstances be traced back to the original individual. | '''De-identification''' removes Private Health Information (PHI) from DICOM files by stripping certain of the DICOM 'tags' from the image headers. There are 1880 DICOM tags but only certain ones need be removed and this can be performed automatically by publicly available software. What remains should include a uniquely assigned study number that allows the image to be linked to other clinical and image data from that same individual. Only the originating institutional investigator holds a code key that could under specially permitted circumstances be traced back to the original individual. | ||
'''Anonymization''' is a higher level of privacy protection beyond de-identification because datasets such as images. though they may contain unique case study identifiers, have nonetheless had all data that allows for trace-back to a given individual removed even as it relates to the originating investigator. | '''Anonymization''' is a higher level of privacy protection beyond de-identification because datasets such as images. though they may contain unique case study identifiers, have nonetheless had all data that allows for trace-back to a given individual removed even as it relates to the originating investigator. | ||
Revision as of 14:11, 26 December 2007
- Return to Main Page Main Page
De-identification removes Private Health Information (PHI) from DICOM files by stripping certain of the DICOM 'tags' from the image headers. There are 1880 DICOM tags but only certain ones need be removed and this can be performed automatically by publicly available software. What remains should include a uniquely assigned study number that allows the image to be linked to other clinical and image data from that same individual. Only the originating institutional investigator holds a code key that could under specially permitted circumstances be traced back to the original individual.
Anonymization is a higher level of privacy protection beyond de-identification because datasets such as images. though they may contain unique case study identifiers, have nonetheless had all data that allows for trace-back to a given individual removed even as it relates to the originating investigator.