Parkinson’s disease (PD) is a common neurodegenerative disorder which has an increasing global burden and no treatment is currently available. Diagnosis of PD is made based on clinical motor manifestations, which occur when a great part of the neurons has already been damaged. Evidence shows that the typical PD motor symptoms can be preceded by a number of abnormalities appearing up to 10 years before motor deficits. Therefore, the correct and precise identification of early neurodegeneration is extremely important for future neuroprotective clinical trials aiming at slowing down or even stopping PD in its early stages.
Among the abnormalities preceding PD motor symptoms, a sleep disorder, known as rapid eye movement (REM) sleep behavior disorder (RBD) is considered the strongest early PD biomarker. Currently, diagnosis of RBD is made by visual analysis of electrophysiological signals recorded during sleep, including electroencephalography (EEG), electrooculography (EOG) and electromyography (EMG). Such analysis is time consuming and prone to subjective interpretation. Because of these limitations, it is difficult to define a large and homogeneous target group for future neuroprotective clinical trials. Automated methods, based on fast, consistent and objective data-driven approaches, have the potential to overcome the time consumption and the subjectiveness of current RBD diagnosis.
This thesis proposes new automated data-driven methods for better, faster and more precise identification of patients with RBD by analysis of sleep electrophysiological signals. First, a comparison of currently available automated methods for RBD detection based on EMG signals is proposed. Such analysis reveals the need of a new algorithm for this purpose. Second, the development and validation of a new and robust data-driven algorithm for RBD identification using EMG signals is presented. The algorithm overcomes previously proposed methods for RBD identification. Finally, the third part presents a fully-automated data-driven method that, based on EEG and EOG signals, can identify RBD and a prodromal stage of RBD.
These new methods have the potential to be used in clinics for faster and more precise identification of patients with early neurodegeneration. The worldwide application of automated methods could guarantee the homogenous identification of a target group for future clinical neuroprotective trials.
To know more about the PhD thesis click here