In September 2019, the Ciklum AI/Deep Learning team attended the 2nd International Congress on Mobile Devices and Seizure Detection in Epilepsy held in Lausanne, Switzerland.
Epilepsy is one of the most common neurological disorders and is characterised by seizures caused by the excessive hypersynchronous activity of neurons in the brain. The main topics addressed at the congress were:
- Innovative technologies for EEG and non-EEG-based seizure detection and prediction (an EEG, or electroencephalogram, is a technique that records the electrical signals emitted by the brain)
- Clinical needs to enhance seizure management
- Reimbursement issues for mHealth epilepsy solutions
The Congress attracted more than 180 attendees from numerous different countries. The attendees included engineers, who are developing brain and heart activity analysis methods to detect and predict epileptic seizures, doctors on the frontline of epilepsy patient treatment and companies working on developments in this field. The Ciklum team delivered poster and oral presentations. The R&D work presented was prepared in partnership with:
- Department of Engineering, University of Palermo, Italy
- Department of Biomedical Engineering and Department of Electronic Engineering, National Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”, Kyiv, Ukraine
- TMO “Psychiatry”, Kyiv, Ukraine
In the poster presentation, entitled “Accessing the information dynamics through EEG source networks in epileptic children”, the team discussed an approach that can achieve good discrimination among different EEG classes and an improved understanding of the physiological mechanisms that underpin the onset of seizures in the pre-ictal phase and the recovery of normal brain function and effective connectivity in the post-ictal phase.
Ciklum’s research engineer, Ivan Kotiuchyi, then delivered an oral presentation entitled “Nonlinear brain-heart interactions in children with focal epilepsy assessed by mutual information of EEG and heart rate variability”.
This presentation described the time-domain analysis our team conducted to better understand the physiological mechanisms that underpin how brain activity evolves before the onset of a seizure and during the recovery phase.
The analysis results suggest that focal seizures are associated with increased brain-heart coupling, which can only be detected by mutual information after a seizure ends. The team concluded that focal epilepsy in childhood is associated with nonlinear brain-heart interaction mechanisms. Thus, linking information on brain-heart dynamics can produce data capable of enhancing our ability to predict seizures.
Looking for a digital partner to realize your deep learning project? Contact Ciklum and we’ll get back to you within 24 hours!