On the Use of t-Distributed Stochastic Neighbor Embedding for Data Visualization and Classification of Individuals with Parkinson's Disease.

Parkinson's disease (PD) is a neurodegenerative disorder that remains incurable. The available treatments for the disorder include pharmacologic therapies and deep brain stimulation (DBS). These approaches may cause distinct side effects and motor responses. This work presents the application of t-distributed stochastic neighbor embedding (t-SNE), which is a machine learning algorithm for nonlinear dimensionality reduction and data visualization, for the problem of discriminating neurologically healthy individuals from those suffering from PD (treated with levodopa and DBS). Furthermore, the assessment of classification methods is presented. Inertial and electromyographic data were collected while the subjects executed a sequence of four motor tasks. The results were focused on the comparison of the classification performance of a support vector machine (SVM) while discriminating two-dimensional feature sets estimated from Principal Component Analysis (PCA), Sammon's mapping, and t-SNE. The results showed visual and statistical differences for all three investigated groups. Classification accuracy for PCA, Sammon's mapping, and t-SNE was, respectively, 73.5%, 78.6%, and 96.9% for the training set and 67.8%, 74.1%, and 76.6% for the test set. The possibility of discriminating healthy individuals from those with PD treated with levodopa and DBS highlights the fact that each treatment method produces distinct motor behavior. The scatter plots resulting from t-SNE could be used in the clinical practice as an objective tool for measuring the discrepancy between normal and abnormal motor behaviors, being thus useful for the adjustment of treatments and the follow-up of the disorder.

Feature visualization and classification for the discrimination between individuals with Parkinson's disease under levodopa and DBS treatments.

BACKGROUND: Over the years, a number of distinct treatments have been adopted for the management of the motor symptoms of Parkinson's disease (PD), including pharmacologic therapies and deep brain stimulation (DBS). Efficacy is most often evaluated by subjective assessments, which are prone to error and dependent on the experience of the examiner. Our goal was to identify an objective means of assessing response to therapy. METHODS: In this study, we employed objective analyses in order to visualize and identify differences between three groups: healthy control (N = 10), subjects with PD treated with DBS (N = 12), and subjects with PD treated with levodopa (N = 16). Subjects were assessed during execution of three dynamic tasks (finger taps, finger to nose, supination and pronation) and a static task (extended arm with no active movement). Measurements were acquired with two pairs of inertial and electromyographic sensors. Feature extraction was applied to estimate the relevant information from the data after which the high-dimensional feature space was reduced to a two-dimensional space using the nonlinear Sammon's map. Non-parametric analysis of variance was employed for the verification of relevant statistical differences among the groups (p < 0.05). In addition, K-fold cross-validation for discriminant analysis based on Gaussian Finite Mixture Modeling was employed for data classification. RESULTS: The results showed visual and statistical differences for all groups and conditions (i.e., static and dynamic tasks). The employed methods were successful for the discrimination of the groups. Classification accuracy was 81 ± 6% (mean ± standard deviation) and 71 ± 8%, for training and test groups respectively. CONCLUSIONS: This research showed the discrimination between healthy and diseased groups conditions. The methods were also able to discriminate individuals with PD treated with DBS and levodopa. These methods enable objective characterization and visualization of features extracted from inertial and electromyographic sensors for different groups.

Estimated features from surface EMG of the lower limb correlate with the subjective sensation of pain

Pain assessment is very important in establishing the efficacy of analgesics and therapies, but because pain is a subjective experience, using methods that represent pain objectively is necessary. A number of biopotentials have been employed in studies of the objective assessment of pain. However, few investigations have considered the peripheral nervous system response to electrical stimulation. The present study evaluated a method for pain quantification based on the analysis of biopotentials. We assessed electromyographic activity that resulted from evoked movements from the nociceptive flexion reflex (NFR). We investigated correlations between stimulus intensity, features extracted from surface electromyography (EMG), and subjective pain reported by subjects using a Visual Analog Scale (VAS). A total of 10 healthy male subjects without any pain disorder, aged 20-27 years, participated in the study. A high correlation (r2 > .87) was found between stimulus intensity and the following features extracted from the EMG: area, root mean square (RMS), and entropy. A high correlation (r2 > .99) was also found between stimulus intensity and subjective pain reported on the VAS. We conclude that estimating features from electromyographic signals that are correlated with subjective pain sensations and the intensity of the electrical stimulus is possible. Entropy, RMS, and the area of the electromyographic signal appear to be relevant parameters in correlations with subjective pain.

Estimated features from surface EMG of the lower limb correlate with the subjective sensation of pain

Pain assessment is very important in establishing the efficacy of analgesics and therapies, but because pain is a subjective experience, using methods that represent pain objectively is necessary. A number of biopotentials have been employed in studies of the objective assessment of pain. However, few investigations have considered the peripheral nervous system response to electrical stimulation. The present study evaluated a method for pain quantification based on the analysis of biopotentials. We assessed electromyographic activity that resulted from evoked movements from the nociceptive flexion reflex (NFR). We investigated correlations between stimulus intensity, features extracted from surface electromyography (EMG), and subjective pain reported by subjects using a Visual Analog Scale (VAS). A total of 10 healthy male subjects without any pain disorder, aged 20-27 years, participated in the study. A high correlation (r2 > .87) was found between stimulus intensity and the following features extracted from the EMG: area, root mean square (RMS), and entropy. A high correlation (r2 > .99) was also found between stimulus intensity and subjective pain reported on the VAS. We conclude that estimating features from electromyographic signals that are correlated with subjective pain sensations and the intensity of the electrical stimulus is possible. Entropy, RMS, and the area of the electromyographic signal appear to be relevant parameters in correlations with subjective pain.(AU)