Generalizable electroencephalographic classification of Parkinson’s Disease using deep learning

Abstract

There is growing interest in using electroencephalography (EEG) and deep learning (DL) to aid in the diagnosis of neurological conditions like Parkinson's Disease (PD). Many existing DL approaches to classify PD from EEG data cite performance metrics in the high 90% accuracies, but may be grossly overestimating their real-word capabilities due to information-leakage between training and testing data. Our aim was to characterize the potential of deep learning for classifying PD using a conservative training approach with unseen external testing data. We used publicly available resting-state EEG data from patients with PD from two seperate centers (University of New Mexico (n = 54) and University of Iowa (n = 28)) for our training and testing sets, respectively. We implemented a channelwise convolutional neural network and tuned it using a subjectwise cross validation approach. We found that an approach commonly cited in the literature overestimated performance in excess of 20%, while our pipeline more conservatively estimated performance by epoch (accuracy: 69.2%; sensitivity: 66.5%; specificity: 72.2%) and by subject (accuracy: 77.4%, sensitivity: 76.9% , specificity: 77.8%). Moreover, we show that our model generalized well to an unseen and external testing dataset without degradation in performance by epoch (accuracy: 77.2; sensitivity: 83.5%; specificity: 71.0%) and by subject (accuracy: 83.8%, sensitivity: 88.6% , specificity: 79.0%). These results highlight the effect of information leakage and serve as a new benchmark for future generalization of DL approaches to classify PD using EEG data.

Distinguishing Parkinson’s Disease with GLCM Features from the Hankelization of EEG Signals

“…Firstly, we achieved an increase of approximately 10% in accuracy compared to previous methods. Some researchers have reported moderate accuracy results such as 88.51% by Shah [39], 69.2% by Sugden [43], 85.4% by Anjum et al [15], 78% by Chaturverdi et al [42], and 88.5% by Aljalal et al [48]. Other researchers reported quite high accuracy results, such as 99.2% by Lee et al [50], 94.1% by Sugden [43], 99.58% and 99.41% by Aljalal [48,51], 94.3% by Vannesta [45], and 99.62% by Yuvaraj [46].…”

“…Some researchers have reported moderate accuracy results such as 88.51% by Shah [39], 69.2% by Sugden [43], 85.4% by Anjum et al [15], 78% by Chaturverdi et al [42], and 88.5% by Aljalal et al [48]. Other researchers reported quite high accuracy results, such as 99.2% by Lee et al [50], 94.1% by Sugden [43], 99.58% and 99.41% by Aljalal [48,51], 94.3% by Vannesta [45], and 99.62% by Yuvaraj [46]. However, some studies in the literature reported have had issues, such as data leakage from the training dataset to the test dataset, unbalanced classes, demographically unmatched groups, or artificially replicated EEG records, leading to very high accuracies.…”

Classifying Parkinson’s Disease Using Resting State Electroencephalogram Signals and U EN -PDNet

“…Their model accomplished 69.2%, 66.5%, and 72.2% of accuracy, sensitivity, and specificity, respectively, epoch-wise performance. Whereas 77.2%, 83.5%, and 71.0% of accuracy, sensitivity, and specificity subject-wise performance [31]. In another attempt, Shaban et al proposed a 20 layered CNN architecture for classifying the PD-on medication from PD-Off medication and the healthy controls.…”