RWE: Analysis Pipeline for Real World Data

Real World Data (RWD) arises from outside controlled clinical studies. The analysis of RWD is hampered by poor data quality (missing data, clinical annotation), data heterogeneity (e.g. EMR, time series, lab data, imaging data) and issues arising from data merging (hidden parameters). These open challenges prohibit the extraction of reliable information with respect to complex, multi-factorial mechanisms for disease development or therapy success. Moreover they significantly hamper the applicability of AI tools such as deep learning.

On the other side, RWD analysis can utilize massive data sets which are not biased by study protocols hence representing the “true” patient populations. Furthermore, it can offer improved comparability/discrimination between heterogeneous sources, enabling reliable interpolations in order to “predict” the behavior of new patients.

The project aims to develop, validate and implement a RWD analysis workflow, predominantly for critical care data from private and public data sources.


Real-World Evidence — What Is It and What Can It Tell Us?

For bachelor/master students, we offer thesis or HiWi positions to assist in the project.

Your profile

  • BSc./MSc. student in Computer Science/Software Engineering/Bioinformatics or other related fields.
  • Good programming skills in Python/R and being comfortable with the Unix-like command line.
  • Good command over English. German is a definite plus. 
  • Prior Biomedical knowledge is preferable but not desired.
  • Motivation to work in an interdisciplinary scientific field across engineering and medicine.