Quantitative Classification of Parkinsonian Gait Disturbances Using Telemetry: Differentiating Akinetic Freezing and Inefficient Tremulous Gait
Abstract
Zhenghua Li and Kenichi Yamamura
Background: We previously demonstrated that freezing of gait (FOG) in Parkinson’s disease (PD) can be objectively assessed using consumer smartwatch telemetry and that FOG episodes introduce significant spatial GPS distance distortion. However, parkinsonian gait impairment comprises distinct motor subtypes, notably complete motor arrest (akinetic freezing) and irregular shuffling gait.
Objective: To determine whether these distinct manifestations can be quantitatively evaluated and differentiated using telemetry metrics derived from structured field walking tests.
Methods: We conducted outdoor walking test trials on a standardized park pathway using Apple Watch Ultra 2 (WorkOutDoors and HealthFit). Quantitative telemetry metrics included the Measurement Error Rate (MER: [(Recorded Distance - Map Distance) / Map Distance] × 100%) and the FOG Ratio (Freezing Time / Total Time).
Results: FOG Ratio strongly correlated with walking speed (r = -0.869, p < 0.001) and cadence (r = -0.848, p < 0.001). Crucially, MER showed no correlation with FOG Ratio (r = 0.135, p = 0.659) or walking speed (r = -0.096, p = 0.755).
Conclusion: FOG Ratio and MER represent orthogonal dimensions of gait impairment. This decoupling establishes a quantitative classification differentiating Akinetic Freezing (high FOG Ratio, low MER) from Inefficient Tremulous Gait (high MER), providing a refined framework for digital gait telemetry.