Objective and automatic classification of Parkinson disease with Leap Motion controller

Abstract

BackgroundThe main objective of this paper is to develop and test the ability of the Leap Motion controller (LMC) to assess the motor dysfunction in patients with Parkinson disease (PwPD) based on the MDS-UPDRSIII exercises. Four exercises (thumb forefinger tapping, hand opening/closing, pronation/supination, postural tremor) were used to evaluate the characteristics described in MDS-UPDRSIII. Clinical ratings according to the MDS/UPDRS-section III items were used as target. For that purpose, 16 participants with PD and 12 healthy people were recruited in Ospedale Cisanello, Pisa, Italy. The participants performed standardized hand movements with camera-based marker. Time and frequency domain features related to velocity, angle, amplitude, and frequency were derived from the LMC data.ResultsDifferent machine learning techniques were used to classify the PD and healthy subjects by comparing the subjective scale given by neurologists against the predicted diagnosis from the machine learning classifiers. Feature selection methods were used to choose the most significant features. Logistic regression (LR), naive Bayes (NB), and support vector machine (SVM) were trained with tenfold cross validation with selected features. The maximum obtained classification accuracy with LR was 70.37%; the average area under the ROC curve (AUC) was 0.831. The obtained classification accuracy with NB was 81.4%, with AUC of 0.811. The obtained classification accuracy with SVM was 74.07%, with AUC of 0.675.ConclusionsResults revealed that the system did not return clinically meaningful data for measuring postural tremor in PwPD. In addition, it showed limited potential to measure the forearm pronation/supination. In contrast, for finger tapping and hand opening/closing, the derived parameters showed statistical and clinical significance. Future studies should continue to validate the LMC as updated versions of the software are developed. The obtained results support the fact that most of the set of selected features contributed significantly to classify the PwPD and healthy subjects.

Data-Driven Models for Objective Grading Improvement of Parkinson’s Disease

“…Indeed, the set of spatiotemporal features that obtained the maximum accuracy included parameters extracted from Gait, Heel-Toe Tapping, Pronation/Supination, Finger Tapping, Arms Swing, and Postural Tremor, confirming the significance of these exercises already found in previous works 4 , 5 for the assessment of motor dysfunction in patients with PD. Furthermore, the selected features are derived from accelerometers and gyroscopes, suggesting that both sensors are equally important for capturing clinically relevant information.…”

“…The chosen feature selection methods have been used in previous studies to obtain the optimum set of parameters to increase the overall accuracy of the system. 4 , 15 …”

“… 22 In this study, we aimed at investigating the relationship between the extracted information from the sensors and the clinical scores provided by the clinicians according to the MDS-UPDRS. To achieve this goal, multiple feature selection methods, 4 , 15 combined with multiple regression models, 11 , 12 , 17 , 25 were investigated to understand which technique could be the most helpful for longitudinal clinical data analysis and long-term monitoring of PD progression. We already investigated classification approach with supervised learning in previous works, considering both binary classification 21 (i.e., PD patients vs healthy controls), and multigroup classification (i.e., PD patients, healthy subjects, and subjects with idiopathic hyposmia that are at risk for developing the disease).…”

A Radiomics Approach to Predicting Parkinson’s Disease by Incorporating Whole-Brain Functional Activity and Gray Matter Structure

“…These results demonstrated the ability of supervised classification methods with a relatively high accuracy. An automatic SVM based study with leap motion controller recruited 16 PD and 12 HCs, and the accuracy was 74.07% with an AUC of 0.675 (Butt et al, 2018). Another study of SPECT imaging using SVM and logistic regression (LR) showed that SVM method produced a higher accuracy of 85% than LR of 83%, and the authors claimed that the SVM-based model could provide better guide for PD stage classification (Hsu et al, 2019).…”

Recognition of American Sign Language Gestures in a Virtual Reality Using Leap Motion

“…Hand gesture recognition is widely researched as it can be applied to different areas such as human-computer interaction [1], robotics [2], computer games [3], education [4], automatic sign-language interpretation [5], decision support for medical diagnosis of motor skills disorders [6], recognition of children with autism [7], home-based rehabilitation [8,9], virtual training [10] and virtual surgery [11]. In industry, gesture recognition can be used in areas requiring very high precision such as to control devices such as robot hands [12] or industrial equipment.…”