Electroencephalography‐based machine learning for cognitive profiling in Parkinson's disease: Preliminary results

Abstract

Background Cognitive symptoms are common in patients with Parkinson's disease. Characterization of a patient's cognitive profile is an essential step toward the identification of predictors of cognitive worsening. Objective The aim of this study was to investigate the use of the combination of resting‐state EEG and data‐mining techniques to build characterization models. Methods Dense EEG data from 118 patients with Parkinson's disease, classified into 5 different groups according to the severity of their cognitive impairments, were considered. Spectral power analysis within 7 frequency bands was performed on the EEG signals. The obtained quantitative EEG features of 100 patients were mined using 2 machine‐learning algorithms to build and train characterization models, namely, support vector machines and k‐nearest neighbors models. The models were then blindly tested on data from 18 patients. Results The overall classification accuracies were 84% and 88% for the support vector machines and k‐nearest algorithms, respectively. The worst classifications were observed for patients from groups with small sample sizes, corresponding to patients with the severe cognitive deficits. Whereas for the remaining groups for whom an accurate diagnosis was required to plan the future healthcare, the classification was very accurate. Conclusion These results suggest that EEG features computed from a daily clinical practice exploration modality in—that it is nonexpensive, available anywhere, and requires minimal cooperation from the patient—can be used as a screening method to identify the severity of cognitive impairment in patients with Parkinson's disease. © 2018 International Parkinson and Movement Disorder Society

Hallucinations, somatic‐functional disorders of PD‐DLB as expressions of thalamic dysfunction

“…Instead, continuous theta activity is observed during REM sleep in PD patients and REM sleep behavior disorder of PD‐RBD . Detection of the theta‐pre‐alpha rhythm in PD is now a core feature of machine‐learning projects developed for assessment of PD and RBD …”

The democratic aspect of machine learning: Limitations and opportunities for Parkinson's disease

“…In this issue, Betrouni and colleagues applied pattern recognition of theta band power in EEG to categorize the degree of cognitive impairment in patients with PD.…”

“…One approach to build classification models using a transformed set of features in much higher dimensions is support vector machine. Prototype methods, such as k‐nearest neighbors, instead reject the idea of building a model and make predictions based on the outcome of similar case examples . The best guess for whether a PD patient has a specific type of cognitive impairment is to see if similar patients (with the same EEG theta power) tend to have the same type of impairment .…”

“…As Betrouni and colleagues point out, data characterizing human phenomena are often high dimensional and heterogeneous and growing in volume as new tools are developed. Thus, they often do not satisfy assumptions required for parametric testing.…”

“…8,13 The best guess for whether a PD patient has a specific type of cognitive impairment is to see if similar patients (with the same EEG theta power) tend to have the same type of impairment. 13 All of these choices have free parameters to fit and require a learning step to optimize their parameters. In the study by Betrouni and colleagues, 13 this was the allocation of PD patients to a specific group of degree of cognitive impairment based on EEG theta power.…”