World Sleep 2022 | AI: predicting disease progression from EEG abnormalities in iRBD

Matteo Cesari, PhD, Medical University of Innsbruck, Innsbruck, Austria, discusses artificial intelligence (AI) in predicting disease progression from EEG scans in individuals with isolated REM sleep behavior disorder (iRBD). Dr Cesari mentions how iRBD is considered one of the strongest biomarkers of early alpha-synucleinopathy– being a precursor to Parkinson’s disease, dementia with Lewy bodies, and multiple system atrophy. Research is currently trying to find biomarkers that allow for greater identification of which iRBD patients will develop disease. It is known that iRBD is a precursor for neurodegeneration – yet a primary research objective is to identify when a patient is likely to manifest disease activity. Dr Cesari describes research currently being conducted on both sleep EEG and resting-state EEG – whereby AI has demonstrated promising results in the resting-state EEG scan. A deep-learning algorithm has also shown the ability of AI in identifying iRBD patients who have later developed Parkinson’s disease – further demonstrating the potential of AI. Simultaneously, there has been work conducted on a deep-learning AI algorithm to identify the conversion of prodromal iRBD to fully-fledged iRBD. When applied to iRBD patients, AI-based analysis can allow for a better understanding of disease development; being non-invasive and potentially cheaper, it is possible that AI could prove to help the development of future clinical trials. This interview took place at the World Sleep Congress 2022 in Rome, Italy.