Objective: In this study, we will investigate the behaviour of Partial Least Squares (PLS) regression for dimension reduction and prediction of motor states of Parkinson’s disease (PD) patients, using upper limb motor data gathered by means of a smartphone.

Background: Design choices related to development of data-driven models significantly impact or degrade their predictive performance. One of the essential steps during development and evaluation of such models is the choice of feature selection and dimension reduction techniques. That is imperative especially in cases dealing with multimodal data gathered from different sources.

Method: Nineteen advanced PD patients and 22 healthy controls were recruited in a single dose, single center study.  The patients received 150% of their individual levodopa-carbidopa equivalent morning dose. All participants repeatedly performed standardized motor tasks according to Unified PD Rating Scale (UPDRS) including UPDRS #23 (finger tapping), UPDRS #25 (rapid alternating movements of hands), and UPDRS #31 (bradykinesia). The patients were video recorded and the videos were observed by 3 movement disorder specialists who rated the UPDRS items, dyskinesia, and Treatment Response Scale (TRS). On each test occasion, the patients performed upper limb motor tasks including tapping and spiral drawing. The smartphone data was analysed and 37 quantitative parameters were extracted. The parameters were used in a PLS regression method to be mapped to mean ratings of the TRS.

Results: The results in terms of correlations between smartphone-based and clinician-derived scores were compared to a previous study using the same data where stepwise regression and support vector machines were used. The correlation coefficients were 0.75 to mean TRS, 0.7 to UPDRS #31, 0.63 to dyskinesia, and 0.48 to the sum of the 3 UPDRS items.

Conclusion: The results from this study show that using PLS is superior than the previous methodology using stepwise regression and support vector machines in terms of prediction performance and provides an objective characterization of motor states in PD.

References: Aghanavesi, S. , Nyholm, D. , Senek, M. , Bergquist, F. & Memedi, M. (2017). A smartphone-based system to quantify dexterity in Parkinson’s disease patients. Informatics in Medicine Unlocked.

« Back to 2019 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/objective-assessment-of-motor-states-in-parkinsons-disease-using-partial-least-squares-and-smartphone-data/