JMIR Publications

ABSTRACT

Background:

Collaborative privacy-preserving training methods allow the integration of locally stored private data sets into machine learning approaches while ensuring confidentiality and non-disclosure.

Objective:

In the present work we assess the performance of a state-of-the-art neural network (NN) approach for the detection of protected health information in texts trained in a collaborative privacy-preserving way.

Methods:

The training adopts distributed selective stochastic gradient descent, i.e. it works by exchanging local learning results achieved on private data sets.

Results:

5 networks trained on separated real-world clinical data sets by utilising the privacy-protecting protocol reach a mean F1 value of 0.955. The gold standard centralised training that is based on the union of all sets and does not take data security into consideration reaches a final value of 0.962.

Conclusions:

Thus, using real-world clinical data our study shows that detection of protected health information can be secured by collaborative privacy-preserving training. In general, the approach shows the feasibility of deep learning on distributed and confidential clinical data while ensuring data protection.