Young Professional Session Invited Talk
Beyond Detection: Towards Personalized and Intelligent Epileptic Seizure Prediction Using AI
- Thursday, November 12, 2026
- 11:00-11:30
Dr. Kiyan Afsari
Lecturer (Assistant Professor I), School of Engineering
University of Wollongong in Dubai.
Dr. Kiyan Afsari received his B.Eng. (Hons.) in Electrical and Electronics Engineering from the University of Wollongong in Dubai (UOWD), UAE, in 2017, his M.Sc. in Robotics from Middlesex University Dubai, UAE, in 2020, and his Ph.D. in Engineering from the University of Wollongong, Australia, in 2026.
He is currently a Lecturer (Assistant Professor I) in Mechatronics at UOWD. His research interests include robotics, autonomous systems, artificial intelligence, machine learning, intelligent systems, biomedical signal processing, and epileptic seizure prediction, with a focus on developing intelligent and adaptive technologies for real-world applications.
He received the Global Excellence in Teaching Award from UOW Global Enterprises in 2025, as well as the Teaching Innovation Award and the Innovation Across Disciplines Award from UOWD in 2026.
Abstract
This talk explores the application of artificial intelligence and deep learning to personalized epileptic seizure prediction using electroencephalography (EEG) signals.
Drawing on Dr. Afsari’s Ph.D. research, it addresses key challenges in developing reliable seizure prediction systems, including inter-patient variability in brain signals, limited patient-specific data, and the computational demands of deep learning models.
The talk introduces a similarity-driven personalization approach using Dynamic Time Warping (DTW), combined with deep learning techniques and EEG channel optimization to improve prediction performance while reducing computational complexity.
Key findings will be presented, including prediction accuracy, sensitivity, performance across different forecasting horizons, and the potential for reduced-channel EEG systems.
Finally, the talk will discuss future directions towards lightweight, adaptive, and wearable AI-powered healthcare technologies that can bring personalized seizure prediction closer to real-world applications.