Assessment of individual risk of dementia in epilepsy:
multimodal brain-based precision prognostics
Project lead: Ira Haraldsen
Institution: Oslo University Hospital (OUS)
Partners: Aalto University, Complutense University of Madrid, Helsinki University Hospital, OsloMET and Università Cattolica del Sacro Cuore-Roma
Funding: South-Eastern Norway Regional Health Authority (HSØ)
Despite years of effort, researchers have made little progress identifying and treating neurological disorders like dementia and other neurodevelopmental diseases. Dr. Ira Haraldsen at Oslo University Hospital and her European partners in the AIRDEM project think this is because we are looking in the wrong direction. They think we might find answers in the indirect mechanisms of diseases like predispositions that start early in life. AIRDEM’s current focus is on developing methods to measure synchronicity – patterns of coordinated neuron activity in the brain – using electroencephalogram (EEG) to test whether these patterns might be precursors to neural degeneration.
Haraldsen was inspired to test this hypothesis during her years studying puberty as a biological phenomenon. In her work, she saw that harmonic, decentralised, playful activity between communicating brain areas was considered healthy. Could a change or disruption of this synchronisation into hyper-synchronicity be an early indicator of pathology? Haraldsen describes the phenomenon as an air traffic controller watching flight patterns: if they see all the airplanes changing course, going in one direction and avoiding another direction, it’s a pretty good indication that there is a problem along the old flight path.
Measuring these patterns is best done with EEG, an inexpensive and well understood technology that measures electrical activity in the brain. However, the math required to map the neural network from EEG and test AIRDEM’s hypothesis is very complicated. To overcome this challenge that has foiled many other labs, AIRDEM has hired an engineer with artificial intelligence (AI) experience and are looking for collaborators with machine learning and data analysis expertise. They are also building a partnership with a private company that is using an AI algorithm to detect cancer that might be able to detect patterns in brain activity as well.
This brain map is needed because physicians currently don’t have a good way to screen at-risk populations before they show symptoms. Ideally, AIRDEM would test people before and after they show symptoms to look for changes in synchronicity, but most patients only come to the hospital once they notice something wrong. While a few hundred healthy people are ultimately needed to develop their brain map, AIRDEM has found a suitable starting population in a group of people with epilepsy, mild cognitive impairment (MCI), and movement disorders.
With these groups of participants AIRDEM is starting their research. They plan to measure EEG patterns in each patient and correlate it with magnetic resonance imaging (MRI), positron emission tomography (PET) scan, and neuropsychological and behavioural testing to create a complete functional brain map. Like the air traffic controller watching planes, the researchers are looking for patterns across time and location to identify signs of potential cognitive decline risk.
Their goal is to develop this measurement algorithm into a complete package that integrates with EEG technology. Physicians in the hospital could apply the algorithm to existing EEG data to quickly screen patients for hyper-synchronicities and start treatment. While the researchers still have a long way to go before they have an answer to their hypothesis, an inexpensive diagnostic tool based on a new understanding of the brain, could have a major positive impact on the healthcare field.