Epileptiform discharges on the electroencephalogram (EEG) reflect an increased risk of epileptic seizures and are assessed visually by default. This is time-consuming, human errors are common, and trained medical experts are not always available. This motivates the development of automatic analyzes for the interpretation of the EEG. We want to automate the detection of epileptiform discharges with unsupervised learning. This makes it possible to "learn" from very large amounts of data, already available in the clinic. We expect to use this to develop AI techniques that can assess EEGs at least as well as medical experts, making epilepsy diagnostics more efficient, cheaper and available 24/7.
- The junior researcher will work on unsupervised anomaly (outlier detection) methods for detecting epileptiform discharges in EEG signals.
- Developing detecting epileptiform discharges software using Matlab and Python.
- Writing academic publications.
- Participation in the supervision of master students.
You will work in an interdisciplinary research team between MST hospital and academic staff at University Twente (Prof dr ir Michel van Putten, Dr. Maryam Amir Haeri, Dr. Stephanie van den Berg) in the fields of neurophysiology, machine learning and statistics. We encourage a high degree of responsibility and independence, while collaborating with close colleagues, researchers and physicians.