Health Enabling Technologies and Data

Reliability of data analytics for low-quality recordings of diverse sensors

Carl-Friedrich-Gauß-Faculty of TU Braunschweig (TU-BS)

The use of health enabling technologies (HET) at home, in the vehicle or worn on the human body generates signal measurements and image data with poor signal quality, shifting offsets and recording gaps. The data quality is much lower as compared to clinical data, but a big data volume is recorded continuously and needs real-time analysis to predict or alarm adverse events. Hence, HET yields novel challenges in signal processing and data analytics, for instance, reliable collection and semantic integration of this data with the electronic health records of the subjects. In addition, the automatic understanding of measurements requires robust algorithms for analysis, such as deep learning.

Learning Objectives: The module covers HET data management from its creation to recording to storing to analysis. By using an open software program, these methods are also applied practically. Furthermore, the basics of semantic interoperability, the determination of a reliable ground truth for the evaluation of algorithms and the model-based methodology for real-time monitoring, event prediction, and emergency detection are covered.

Target Group: The target group is master’s students of computer science, medical informatics, and related study courses. In this module, students learn to understand and solve challenges and difficulties in biomedical signal processing of HET-data and their integration into health records. Students also know and understand the basics of semantic interoperability. Besides, students can actively apply and analyze signal and image processing methods, both in theory and in practice with Python and R. Students learn how to deal with linear and non-linear noise and to evaluate the suitability of methods for processing or preprocessing HET-data as well as to work out proposals for suitable methods. The module is to be embedded in the master's program in computer science, business informatics, and comparable master's programs in the elective field of medical informatics.

Embedding: The module is to be integrated into the curriculum of the Carl-Friedrich-Gauß-Faculty of TU Braunschweig.