Saccadic Variability in Patients with Parkinson's Disease

Quantifying temporal irregularities in saccadic movements using Dynamic Time Warping for clinical interpretation.

Authors Kyoungwon Baik*, Seoyeon Byun*, Seooui Kwag, Hyunsoh Park, Yukang Kim, Tonghoon Woo, Sun-Uk Lee, Nyoung Lee, Gerard J. Kim, Byung-Jo Kim, Ji-Soo Kim, Kun-Woo Park
My Role Co-First Author (Equal Contribution), Time-series Analysis Pipeline Implementation
Project Period Aug 2024 - Nov 2025

Background and Objectives

Patients with Parkinson’s disease (PD) are known to exhibit delayed saccades and saccadic hypometria, but the diagnostic yield of saccadic parameters is limited in the clinical setting. We hypothesized that, as motor fluctuations, fluctuation would also be documented on saccadic parameters, which would be feasible for interpretation and clinical application.

Methods

This cross-sectional study is based on a 114 prospectively recruited de-novo PD cohort in South Korea, collected from February 2022 to August 2024. All participants conducted horizontal and vertical saccades following a target that alternately appeared on a light bar every 2 s. We analyzed the mean, SD, and coefficient of variance (CV) of saccadic latency, amplitude, and velocity. Dynamic time warping (DTW) was applied to each saccadic parameter to assess the temporal variability of saccadic sequences, including control-deviation DTW (CD-DTW) and population-variability DTW (PV-DTW). The data were compared to those obtained from 70 age- and sex-matched healthy participants.

This video shows researchers testing the experimental equipment.
This video shows researchers testing the experimental equipment.

My Role

I quantified temporal irregularities that conventional summary statistics could not capture by implementing a Dynamic Time Warping-based time-series analysis pipeline.

Outputs

This project has resulted in a manuscript that is currently under review.