As a result of the aging population, an increasing number of patients suffer from cardiac arrhythmias. This makes long-term and high-quality ECG recording highly desirable. In close collaboration with the Bern University of Applied Sciences and the Inselspital Bern, our group is investigating an innovative esophageal ECG recorder that may improve state-of-the-art diagnostics. A key challenge of this project is to miniaturize and integrate the recording electronics into an implantable tube. The massive amount of data generated during up to 30 days of recording need to be compressed before storage. Conventional algorithms fail since they are associated with high computational effort, increasing total power consumption, i.e. the battery size. Nevertheless, the ECG signal is sparse in the time domain. Non-uniform sampling based on signal level crossing makes use of this fact. Although this novel approach does not conform to the Nyquist sampling theorem, full ECG signal reconstruction is possible using much fewer samples than in classical approaches. Memory and battery space requirements can likewise be saved.
Project members: Andreas Häberlin, MD, PhD, Thomas Niederhauser, PhD, Thanks Marisa, MSc