Neurosciences and Mental Wellbeing
Machine learning for physiological monitoring and epilepsy diagnosis
Supervised by Dr Levin Kuhlmann
A GRIP project involving Alfred Health and Epworth Eastern, this initiative will involve monitoring epilepsy patients and using machine learning to improve the speed and accuracy of diagnosis. Conducted in two stages, the study will first assess current techniques for detecting epileptiform activity and develop new ones to produce more precise results. Afterwards, the researchers will use long-term data to create novel seizure detection/prediction methods.
Personalised Seizure Prediction
Kuhlmann, L., Cook, M., Freestone, D. R., D'Souza, W. & Burkitt, A. N.
National Health & Medical Research Council (NHMRC)
State-of-the-art personalised seizure prediction algorithms will be developed to alleviate the stress associated with the unpredictability of epileptic seizures. These algorithms will be designed to run on portable or implanted devices that record brain activity and look for specific signal changes that are predictive of an individual patient’s seizures. Warnings would then be given to patients so they can move to safety or a device could be activated to prevent seizures from occurring.