Using electroencephalography to predict development of Parkinson’s Disease

Parkinsons Disease (PD) is a neurodegenerative disorder caused by the death of dopamine-containing neurons. Common symptoms include hand tremors, impaired movement, balance, and cognition. The loss of neurons in the midbrain causes significant changes in cortical and subcortical activity that can be detected by techniques such as Electroencephalography (EEG). In previous works, EEG recordings have been used to identify the onset of Parkinsons via a technique called Linear predictive coding EEG algorithm for Parkinson’s disease (LEAPD).

This Master’s work extends the usage of LEAPD to several new classes of PD subjects to test the effectiveness of identifying other common complications involved with Parkinson’s disease. LEAPD is first used to identify depression in patients with PD, and PD in patients with depression. EEG recordings were then truncated to ensure that the classification was robust. Then, its accuracy is measured in identifying future mortality in patients. Correlation of LEAPD with a normalized future Parkinson’s Disease Composite Scale (PDCS) score is checked. Finally, recordings of patients using Deep Brain Stimulation are analyzed as a method of identifying Unified Parkinsons Disease Rating Scale (UPDRS). The correlation between LEAPD and Universal Parkinsons Disease Rating Scale (UPDRS) is then tested. The best performing channels in each case are further analyzed to prove their effectiveness.

Overall, this work finds that LEAPD is highly capable of discriminating between different classes of individuals in each scenario. The final section in this work discusses future applications of LEAPD and further areas of study.