Secure computation

In the MHMD system, data are locally stored at mutually distrustfully parties (i.e., hospitals) forming a distributed database. As an alternative to collect all data in a central location to perform statistical computations, the Consortium has developed the approach of “bringing algorithms to the data”.

In other words, the MHMD system allows to perform a distributed data mining computation accessing data at each location of the distributed database and running the algorithm at local level, without physically transferring the data and sending back computation results only.

Secure multiparty computation (SMC)

SMC is a subfield of cryptography with the goal to create methods for parties to jointly compute a function over their inputs, keeping these inputs private. SMC allows a set of distrustful parties to perform the computations in a distributed manner, while each of them alone remains oblivious to the input data and the intermediate results. The computation is considered secure if, at the end, no party knows anything except its own input and the results.

Our Consortium has been constructing a toolkit of secure computations to build data mining algorithms where only aggregated sufficient statistics (with provable bounds on the information released) is collected from local hospital nodes.



Homomorphic encryption (HE)

The property of an encryption scheme that allows for performing operations on encrypted data is called homomorphism. HE represents an encryption scheme allowing for computations on encrypted data, where data is encrypted before being sent to the computing service, and computations are performed on encrypted data. Once the results are available, they are sent back and decrypted at the source. The computing service has access only to the encrypted data, and since the decryption key is not available to the service, no personal or useful information can be extracted.


To know more: 

  • Vizitiu, A., Niţă, C. I., Puiu, A., Suciu, C., & Itu, L. M. (2019, June). Towards Privacy-Preserving Deep Learning-based Medical Imaging Applications. In 2019 IEEE International Symposium on Medical Measurements and Applications (MeMeA) (pp. 1-6). IEEE.
  • Vizitiu, A., Nitǎ, C. I., Puiu, A., Suciu, C., & Itu, L. M. (2019, July). Privacy-Preserving Artificial Intelligence: Application to Precision Medicine. In 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) (pp. 6498-6504). IEEE.