Objective: To develop a sensor-based pipeline to identify lower limb dystonic features in X-linked dystonia parkinsonism (XDP).

Background: XDP is a rare neurogenetic combined movement disorder involving dystonic and parkinsonian features. Limb dystonia severity is notoriously difficult to rate using clinical rating scales, given semiological complexity and dynamic features, while sensor-based motion analysis may provide objective biomarkers.

Method: We assessed 28 XDP patients, 4 carriers, and 4 controls using 17 inertial measurement unit sensors during a standardized examination, including rest and active lower limb tasks. Clinical dystonia severity was assessed with the Burke-Fahn-Marsden Dystonia Rating Scale (BFM). Dystonia semiology and directionality during individual tasks were documented and used as labels. Feature selection was employed to identify dystonic markers showing the greatest median separation between dystonic and non-dystonic populations. Features were extracted, standardized and categorized by dystonia severity (BFM).

Results: During resting tasks, dystonic foot dorsiflexion was easier to distinguish from non-dystonic participants than dystonic foot inversion. During active tasks, peak foot velocity emerged as a robust, semiology-independent measure distinguishing dystonia severity across tasks and outperformed joint angle measures, with reliability across semiologies. Feature selection reduced an initial 504 kinematic parameters to 69 critical markers. Ankle dorsiflexion/plantarflexion and eversion/inversion joint angles were highly discriminative during seated active and resting conditions. Sensor-based analyses showed limitations, particularly in detecting resting dystonia involving complex, multidirectional movements, where only a subset of joint angles differentiated dystonia severity. There was also variability in sensor sensitivity in capturing certain common dystonia semiologies, such as toe movements, given the use of a single dorsal foot sensor.

Conclusion: These findings underscore the capability of wearable sensor-based analysis to identify kinematic markers of lower extremity dystonia in XDP, with widespread applicability to other forms of dystonia. Despite limitations to this approach, sensor-based models demonstrate strong potential as objective, rater-independent tools to enhance clinical assessment and monitoring of dystonia.

« Back to 2025 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/wearable-sensor-based-assessment-of-lower-extremity-dystonia-in-x-linked-dystonia-parkinsonism/