Systems and methods for computing with private healthcare data
Techniques are provided for computing with private healthcare data. The techniques include a de-identification method including receiving a text sequence; providing the text sequence to a plurality of entity tagging models, each of the plurality of entity tagging models being trained to tag one or more portions of the text sequence having a corresponding entity type; tagging one or more entities in the text sequence using the plurality of entity tagging models; and obfuscating each entity among the one or more tagged entities by replacing the entity with a surrogate, the surrogate being selected based on one or more attributes of the entity and maintaining characteristics similar to the entity being replaced.
1. A de-identification method comprising:
receiving a text sequence;
providing the text sequence to a plurality of entity tagging models, each of the plurality of entity tagging models being trained to tag one or more portions of the text sequence having a corresponding entity type;
tagging one or more entities in the text sequence using the plurality of entity tagging models;
aggregating tagged entities from the text sequence identified by the plurality of entity tagging models;
passing the aggregated tagged entities through one or more dreg filters, each of the one or more dreg filters being configured to filter a corresponding entity type based on at least one of a rule-based template or a pattern matching filter; and
obfuscating each entity among the one or more tagged entities by replacing the entity with a surrogate, the surrogate being selected based on one or more attributes of the entity and maintaining characteristics similar to the entity being replaced.
2. The de-identification method of claim 1 , wherein obfuscating each entity among the one or more tagged entities comprises replacing two or more entities that refer to a common subject with a common surrogate.
3. The de-identification method of claim 2 , wherein the common surrogate is selected based one or more attributes of the two or more entities.
4. The de-identification method of claim 2 , wherein the common surrogate is selected based on a gender associated with the two or more entities.
5. The de-identification method of claim 2 , wherein the common surrogate is selected based on an ethnicity associated with the two or more entities.
6. The de-identification method of claim 1 , wherein the tagging one or more entities comprises tagging two or more personal names; and
the obfuscating each entity among the one or more tagged entities comprises replacing each of the two or more personal names with a different surrogate.
7. The de-identification method of claim 1 , wherein the obfuscating each entity among the one or more tagged entities comprises replacing two or more entities that refer to a common person with surrogates that match a gender associated with the common person.
8. The de-identification method of claim 1 , wherein the obfuscating each entity among the one or more tagged entities comprises replacing two or more entities that refer to a common person with surrogates that match an ethnicity associated with the common person.
9. The de-identification method of claim 1 ,
wherein the obfuscating each entity among the one or more tagged entities comprises replacing two or more tagged entities that represent dates with surrogate dates, and
wherein the surrogate dates are based on the two or more tagged entities altered by a random value.
10. The de-identification method of claim 9 , wherein dates associated with a common patient are altered by the same random value.
11. The de-identification method of claim 1 , wherein the obfuscating each entity among the one or more tagged entities comprises scrambling two or more entities that represent numeric identifiers with random values to scramble the numeric identifiers.
12. The de-identification method of claim 1 , wherein the text sequence comprises at least a portion of an electronic health record.
13. The de-identification method of claim 1 , wherein at least one of the plurality of entity tagging models is trained to tag entities of an entity type, the entity type including at least one of a personal name, an organization name, an age, a date, a time, a phone number, a pager number, a clinical identification number, an email address, an IP address, a web URL, a vehicle number, a physical address, a zip code, a social security number, or a date of birth.
14. The de-identification method of claim 1 , wherein at least one of the plurality of entity tagging models tags entities based on a rule-based algorithm.
15. The de-identification method of claim 1 , wherein at least one of the plurality of entity tagging models includes a machine learning model based on learning from sequences of text.
16. The de-identification method of claim 1 , further comprising whitelisting one or more portions of the text sequence, wherein the one or more whitelisted portions are not provided to the plurality of entity tagging models.
17. The de-identification method of claim 1 , each of the plurality of entity tagging models is trained to tag one or more portions of the text sequence to achieve a performance metric above a predetermined threshold.