Objective: To determine whether markerless motion tracking during activities of daily living (ADLs) can identify kinematic changes during medication wearing-off as Parkinson’s disease (PD) patients transition between natural daily maximum (ON) and minimum (OFF) levodopa states.
Background: Motor symptom fluctuations over the medication cycle are a hallmark of PD, and medication regimens aim to minimise them. However, these are typically tracked subjectively using daily diaries or wearable sensors that capture tremor or slowness during instructed tasks, but are less informative during free behaviour. Capturing true medication and wearing-off effects on patients’ daily living requires a different set of tools.
Method: Kinematics from 38 mild-to-moderate PD patients performing two ADLs (making tea and making toast) were recorded using three Azure Kinect sensors during their OFF (pre-dose) and ON (one hour post-dose) states. Motion features were extracted from the participants’ skeletons per frame, including spine curve, shoulder slope, centre-of-mass sway, ankle distance, and wrist-body distance. Distributional statistics were calculated for each feature and compared between the ON and OFF states with univariate Wilcoxon signed-rank tests. Principal Component Analysis (PCA) was then applied and paired PC score differences were evaluated.
Results: Univariate analysis revealed task-specific sensitivities. Shoulder slope range, spine curve range, and ankle distance range significantly increased in the ON state in both ADLs. When combined in PCA, the natural levodopa fluctuation significantly altered the first two PCs, highlighting that more variance in the data can be explained by fluctuations than by individual differences between participants.
Conclusion: Natural daily fluctuations during medication wearing-off produce robust, quantifiable kinematic changes across ADLs. This demonstrates that markerless motion tracking can provide an ecologically valid approach for capturing real-world treatment efficacy in Parkinson’s disease.
To cite this abstract in AMA style:
J. Yun, S. Haar. Markerless Kinematics of Levodopa Response During Activities of Daily Living in Parkinson’s Disease [abstract]. Mov Disord. 2026; 41 (suppl 1). https://www.mdsabstracts.org/abstract/markerless-kinematics-of-levodopa-response-during-activities-of-daily-living-in-parkinsons-disease/. Accessed October 1, 2026.« Back to 2026 International Congress
MDS Abstracts - https://www.mdsabstracts.org/abstract/markerless-kinematics-of-levodopa-response-during-activities-of-daily-living-in-parkinsons-disease/