Header Image

  About   Projects   Teaching   Publications   Team   Outreach


Signal processing of cardiovascular data

The core of my work is focussed on extracting information from biosignals acuired in clinical practice. During my PhD I developed msPE which is a robust framework for computing parameters of Gaussian-related functions. You can find source code on Github and details in our IEEE TSP paper. See our IEEE JBHI paper for details on the application of msPE to ECG signals from a clinical database. I am proud for receiving the Johann Peter Süßmilch-Medaille for the latter work.

Signal processing in IoT healthcare applications

During my time at the Peter L. Reichertz Institute for Medical Informatics, I worked on using signal processing methods in different IoT applications. This includes an 5G edge platform for ECG real-time analysis (IEEE EMBC 2021), sensor fusion of multiple biosignals for heart rate assessment within a smart car (IEEE EMBC 2021), and sleep analysis using smart textiles (CDBME 2021).

Signal processing in smart home healthcare applications

During my time at the Peter L. Reichertz Institute for Medical Informatics, I worked on signal processing techniques for smart home data to enable applications addressing the challenges of an aging society. See our Sensors paper for an overview of the state-of-the-art regarding sensors application in smart homes. Furthermore, I dived a little into video-based fall (SPIE MI 2021) and activity recognitation (SPIE MI 2022) and prepared a study on capacitive ECG embedded into chairs (PLOS One).

Signal processing in Magnetic Resonance Imaging

During my PhD, I developed methods for cardiac triggering/gating based on photoplethsymography imaging in the context of ultra-high-field magnetic resonance imaging. See my ISMRM 2017 slides and ISMRM 2018 poster for the idea and application, respectively. Furthermore, I worked on ECG signals disturbed by high-field artifacts (CDBME 2019).

Medical Knowledge representation

I dived a little into how to represent (medical) knowledge. See the MIE 2021 and MEDINFO 2021 papers for details on how to use SHACL for detecting errors in registries or for representing medication error reports, respectively. Furthermore, as part of a BMBF project we started developing a digital decision- and workflow-support system drafted in our Therapeutic Advances in Chronic Disease publiation.

Accident & Emergency Informatics

Moreover, I was involved in establishing the research field of Accident & emergency informatics at the Peter L. Reichertz Institute for Medical Informatics (2020-2022), including the ISAN project. See the freely-available E-Book on that topic to get a broad overview of that interesting field of research and also see our Methods Inf Med paper which introduces basic concepts of a unique identifier for accidents.

Crowdsourcing in medical imaging

Machine learning is a promising approach in medical image processing but large amounts of training data are required and expert annotations are costly. We work towards establishing crowdsourcing as a cost-efficient alternative which also provides knowledge transfer.