Objective: To predict global motor disabilities using symptomatology assessed with inertial sensors, medication regimen, and cognitive deficits through: (1) a clustering method on performance metrics to assess different motor tasks in their capacity to reveal distinctive performance groups; and (2) a multinomial regression model to predict the performance group allocation based on the patient’s condition.

Background: Advances in mobile technologies allow an objective appraisal of the state of the patient, but the sheer amount of data produced needs to be carefully examined to ensure that the results are clinically meaningful.

Method: 107 patients with PD equipped with a full-body inertial measurement system performed different tasks such as eating soup, completing a timed up and go (TUG) and performing the Purdue Pegboard test. Performance was assessed using the time required to complete the task, number of errors and number of pins placed. K-means method was used for clustering the overall performance. Participants also underwent an evaluation of symptoms present. Change point analysis was used to determine at which point a specific symptom modified the group performance. A multinomial regression model was then derived using 80% of the data, and validated with the remaining 20%.

Results: Clustering delineated four distinct performance groups: normal behaviour, slightly affected in fine motor tasks, affected only in TUG, and affected in all areas. The statistical model revealed that low-level dyskinesia increased the likelihood of being in the normal group. Increased postural instability increased the risk of being in the group affected in TUG only. Finally, the presence of Amantadine as part of the medication regimen appears to decrease the likelihood of being part of the group affected in TUG only or fine motor task.

Conclusion: This innovative approach takes advantage of the resolution obtained with inertial sensors, while addressing the negative impact of having continuous and often non-linear measurements, hence increasing the chance to get meaningful answers to important clinical questions. As a result, this approach is considered to be a step forward in the development of methods to predict the impact of the disease on a person’s ability to perform daily activities based on clinical evaluation and medication regimen.

« Back to 2019 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/predicting-the-functional-impact-on-the-performance-of-daily-activities-in-people-with-pd-a-novel-approach-based-on-objective-measures-obtained-by-inertial-sensors/