Vision

COPERIMOplus

COronavirus PErsonalized RIsk MOdels

COPERIMOplus is a an attempt to use modern AI technologies to "learn" patterns in clinical progression of COVID-19 and to use the resulting models for personalized risk assessment.

Of course, such an AI approach will need a lot of high-quality, patient-level data. We are fully aware of the fact that sharing patient-level data raises issues related to data privacy and data autonomy.

We have therefore established partnerships with leading institutions across the globe. We re-use data from various types of published studies and cooperate closely with data owners in China, Korea, Europe and the US to improve our models.

As the sharing of patient-level data imposes a major challenge for many data (study) owners, we have established a strong legal framework and implemented data protection measures to guarantee data privacy and data protection. Moreover, we have created technical solutions that allow us to "bring the algorithms to the data" – a strategy that allows for re-training of pre-trained models without compromising the privacy of patient-data.

 

NEWS: Press release from the Universitätsklinikum Erlangen (in German):
"Hoffnung für schwerst kranke COVID-19-Patienten"

This research was done in the "COPERIMOplus" initiative and supported by the Fraunhofer "Internal Programs" under Grant No. Anti-Corona 840266.