Aris Gkoulalas-divanis  Aris Gkoulalas-divanis photo       

contact information

Technical Lead on Data Protection and Privacy, Watson Platform for Health
Cambridge Research Center, Cambridge, MA USA
  +1dash781dash346dash7477

links

Professional Associations

Professional Associations:  ACM  |  ACM SIGKDD  |  BDVA  |  British Computer Society  |  IEEE   |  IEEE Computer Society  |  Sigma Xi


Research work: G. Loukides, A. Gkoulalas-Divanis, B. Malin. Anonymization of Clinical Profiles for Validating Genome-Wide Association Studies. Proceedings of the National Academy of Sciences of the United States of America (PNAS), 2010. Impact factor: 9.83.

Contribution: Our PNAS paper proposes the first approach that effectively guarantees protection of patients' privacy in published patient-specific genetic and clinical data, while maintaining researchers' ability to analyze the data. The current US standard for medical indexing, the International Statistical Classification of Diseases and Related Health Problems, works by assigning a unique number to a health condition. Patient privacy can be threatened when personal information is linked to genetic information using these codes that are available through public databases and medical records. In this paper, we developed a method that replaces single ICD codes with a minimal set of semantically related codes. Based on this modification strategy, the method allows clinical information to be released in a way that preserves privacy, while providing the medical and genetic connections that researchers rely upon for Genome-wide Association Studies. The proposed algorithm was tested against simulated malicious attacks using actual information about more than 2600 patients, assuming a potential hacker knew a patient's identity, some or all of a patient’s ICD codes, and whether a patient's record is included in the released data. The technique successfully foiled all attempts to uncover a patient’s private information and retained useful information for validating Genome-Wide Association Studies.

Media coverage: This work has attracted attention from the media in the US, leading to several interviews and articles in the press: Computer Weekly, GenomeWebNature News, Science NewsScientific American, Technology ReviewThe RegisterWiredZDNet, KDNuggets, bNET industry report, Vanderbilt Reporter.