Format

Send to

Choose Destination
Comput Biol Med. 2019 Feb;105:144-150. doi: 10.1016/j.compbiomed.2018.11.018. Epub 2018 Dec 24.

Private naive bayes classification of personal biomedical data: Application in cancer data analysis.

Author information

1
Department of Computer Science, The Graduate Center, CUNY, New York, NY, 10016, USA; The Department of Computational Medicine and Bioinformatics, University of Michigan Center for Integrative Research in Critical Care (MCIRCC), Ann Arbor, MI, 48109, USA. Electronic address: anwood@med.umich.edu.
2
Department of Mathematics, The Graduate Center, CUNY, New York, NY, 10016, USA; Department of Mathematics, The City College of New York, New York, NY, 10031, USA.
3
The Department of Computational Medicine and Bioinformatics, University of Michigan Center for Integrative Research in Critical Care (MCIRCC), Ann Arbor, MI, 48109, USA; Emergency Medicine Department, University of Michigan, Ann Arbor, MI, 48109, USA.
4
Department of Computer Science, The Graduate Center, CUNY, New York, NY, 10016, USA; Department of Computer Science, Tandon School of Engineering, New York University, New York, NY, USA.

Abstract

Clinicians would benefit from access to predictive models for diagnosis, such as classification of tumors as malignant or benign, without compromising patients' privacy. In addition, the medical institutions and companies who own these medical information systems wish to keep their models private when in use by outside parties. Fully homomorphic encryption (FHE) enables computation over encrypted medical data while ensuring data privacy. In this paper we use private-key fully homomorphic encryption to design a cryptographic protocol for private Naive Bayes classification. This protocol allows a data owner to privately classify his or her information without direct access to the learned model. We apply this protocol to the task of privacy-preserving classification of breast cancer data as benign or malignant. Our results show that private-key fully homomorphic encryption is able to provide fast and accurate results for privacy-preserving medical classification.

KEYWORDS:

Data privacy; Fully homomorphic encryption; Medical information systems; Predictive models; cryptographic protocols

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center