Virtual patient cohorts: breaking the data deadlock

September 4, 2020 2:30 - 4:00

Location: Virtual Room 2


Data is the key to unlocking unparalleled medical revolutions, and yet also a major roadblock for translational research. Patient data is difficult to access, and even more difficult to use. Data silos and lack of interoperability hinder the use of data in collaborative research.
While researchers need real-world evidence to support their findings in controlled studies, the legal safeguards of the General Data Protection Regulation have added further data processing obligations to the already lengthy compliance burden.
How can we respond to the challenge of using sensitive patient data to support science, without compromising their privacy? What if we could create artificial, as-good-as-real data sets, and integrate different data sources for risk-free research? The synthetic data revolution could bring just that.
Machine learning is enabling the virtual simulation of highly realistic patient cohorts. Synthetic data mirrors both the characteristics and diversity of actual patients, but never represents one single real human being. Synthetic data can therefore be freely processed, mixed and matched, analysed and shared without infringing patients’ privacy.
Breaking the data deadlock means increasing the robustness of the data and results, and dramatically speeding up learnings. The application of virtual patient models can have immeasurable impacts on, among others, the clinical part of drug development – the most expensive and lengthy phase of bringing new treatments to patients.
Scientists have proven that it is possible to build a fully functioning virtual patient cohort. We will soon see the simulation of entire trials with a virtual dementia cohort. Next is to foster international efforts to build an open, global cohort of virtual patients to be used by researchers, students and practitioners alike.
Are we ready for it? This session will discuss the tremendous progress made by data scientists in this field, and explore the scope for virtual patient applications.

Go back