Unlocking the power of big data – Without compromising patient privacy

  • Posted on 12.12.2018

Unlocking the power of big data – Without compromising patient privacy

Sorin STIRCU - BioPic

Sorin Stircu

Regional MedTech Strategist at E*HealthLine - Europe Global Office


Deep learning is a form of machine learning with the potential to extract meaningful information from the mountains of data generated by healthcare companies. One of the major obstacles to embracing deep learning in the world of health arises from challenges around data sharing.

As we all know, while there is no shortage of patient data, it may be dispersed across several sites. For example, data on a single patient could be found in their doctor’s office, on hospital systems, in a clinical trial database, a patient registry, company files and held by an insurance company. Given privacy and data protection issues, working with all this data can be very challenging.

However, it is not impossible. This is a problem that I believe can be solved through what is known as federated distributed learning procedure. That’s quite a mouthful but let me explain.

Federated distributed learning is defined as learning from data, without removing the data from the location where it is stored. The beauty of this, in my opinion, is that it offers a chance to unlock the potential of the data without moving it or jeopardising data protection.

Collaborative data mining

It also opens to door to great collaboration. Data collaborative innovation is when a group of actors from different data domains work together toward common goals. Preservation of privacy in data mining has emerged as an absolute prerequisite for exchanging confidential information in terms of data analysis, validation, and publishing.

Our approach to advancing federated deep learning has been fueled by growth in the number of sources of data, including increased access to published research, electronic health records, the various “omics” fields and even social media.

I am building federating deep learning models and algorithms that analyze molecular and medical imaging libraries, as well as patient profiles to uncover complex biomarker patterns behind diseases. Designed to assist researchers in preclinical and clinical studies and clinicians in healthcare settings, it can build models for automated diagnostics, prediction of treatment outcomes or clinical trial optimization. We invite researchers from academia to collaborate with us on joint research projects.

It is my belief, that the implementation of federated deep-learning models will unlock the trust power of data. This can help bring the benefits of deep learning to domains where data owners are precluded from sharing their data by confidentiality concerns. This could be empowered by the big data analysis and deep learning approaches that are currently driving the digital transformation across all healthcare industries.

Increasing AI investments in drug discovery by big pharma companies suggests the medicines industry has not only woken up to, but is actively embracing, the benefits of federated deep learning to identify and screen drugs, more accurately predict drug candidates and cut R&D costs and time. The medical technology industry should take note.

The comments are closed.