Computational neuroscience

Rapid improvement of computational power over the last decades brought the unique opportunity to effectively apply advanced computational methods and models of machine learning on extensive electrophysiological and multimodal data. At the same time, exponential growth of storage capacities allows for automatic processing of prolonged electroencephalographic (EEG) recordings with high sampling rates. Unfortunately, the current clinical gold standard does not fully benefit from these new technologies yet.

Computational neuroscience research group is focused on development of advanced methods for signal processing in neurology, study of the human brain electrophysiology and neurological diseases. Our ultimate aim is open-source and open-science development of advanced technologies and subsequent implementation of these tools into clinical practise in order to improve medical treatment, lower risk and reduce time of patient’s hospitalization.

Our group consists of biomedical engineers and neuroscientists and for many years collaborates with scientists, medical doctors, electrophysiologists and students from Brno University of Technology, Masaryk University, St. Anne’s hospital in Brno and international centers: Mayo Clinic in Minnesota, USA and Montreal Neurological Institute and Hospital in Canada. Currently, our research is mainly focused on patients with pharmacoresistant epilepsy and patients with Parkinson’s disease.

The main research interests are:

Development of methods for automatic processing

  • Development of automatic and semi-automatic tools for data quality assessment and pre-processing of extensive EEG recordings.
  • Development of a multicentric database with prolonged recordings, which should allow for testing on bigger samples across different institutions.

Research and development of analytical methods

  • Methods for broadband EEG signal processing - analysis of interictal epileptic discharges and high frequency oscillations (HFO).
  • Connectivity and mutual interactions between anatomical structures of the human brain, analysis of the epileptogenic zone functional connectivity.

Application of developed methods in neurology

  • The basic research of motor and cognitive processes.
  • Analysis of the epileptogenic zone function, dynamics of epileptic seizures.
  • Effectivity of deep brain stimulation (DBS).
  • Machine learning models:
    • localization of the epileptogenic zone
    • prediction of surgical outcome in epilepsy surgery
    • seizure forecasting and seizure prediction
    • prediction of the effect of vagal nerve stimulation (non-invasive scalp EEG study).
  • Implementation of the developed tools into clinical practise.
  • Therapy: aimed estimation, selective micro ablations.