Research Project , 2020

Although the HRV signal has been studied extensively in the clinical settings, the quality of collected data is still a challenge. For example, the signal’s variability needs to be correctly linked/labelled by patient’s status (e.g. sedation, body movement, cerebral ischema), which is difficult to be done á posteriori.


This project addresses problems with labelling of HRV data with events related to the patient’s cerebral diseases. We plan to map methods and techniques to use for highly reliable labelling of continuously streamed HRV signals from individual patients to the events that happen in the patient’s brain by interpreting the EEG signals.


In this project we conduct a series of participatory observations (sv: auskulera) at the SU:s Neurointensive Care Unit and Clinical Neuroscience. These give us the understanding of how toolchains for collecting, processing and visualizing of ECG and HRV are used in practice for setting diagnoses about Delayed Cerebral Ischema and related diseases. They will also provide us with the understanding how to use federated learning to link EEG signals (and their interpretation) with ECG (and HRV) signals to enable high quality learning.


Miroslaw Staron (contact)

Chalmers, Computer Science and Engineering (Chalmers), Software Engineering (Chalmers)


Sahlgrenska University Hospital

Gothenburg, Sweden


Chalmers AI Research Centre

Funding Chalmers participation during 2020

Related Areas of Advance and Infrastructure

Information and Communication Technology

Areas of Advance

Health Engineering

Areas of Advance

More information

Latest update