Quantifying Slowness in Parkinson Disease Using a Serious Game: Cross-Sectional Study.

Background: Slowness in voluntary movements is a hallmark of Parkinson disease (PD); yet, objective measurement outside clinical settings is limited. Serious games represent a promising alternative to extract motor performance metrics during interactions. However, evidence on the effectiveness of these games in discriminating motor performance between individuals with and those without PD is still scarce. Objective: This study aimed to objectively assess slowness using the RehaBEElitation serious game, based on in-game estimated measurements. Methods: The study included 15 individuals with mild to moderate PD (Hoehn and Yahr I-III), assessed in both ON and OFF medication states, and 15 age- and gender-matched healthy controls (10 men and 5 women; mean 66.27, SD 9.13 years; range 45-82 years). All participants played each phase of the game on the easiest level. Slowness was evaluated by detecting the voluntary movement of the gyroscope signals using the singular spectrum analysis method. The response time (RT) and angular velocity (AV) of the participants while playing RehaBEElitation were estimated. Group-level comparisons were performed to investigate the presence of slowness patterns across conditions. The Kruskal-Wallis test and Wilcoxon signed rank test with Bonferroni correction were used to confirm the differences between groups, and the effect size was estimated using eta square (η2). Spearman correlation analyses were conducted to examine associations between the RT and AV and the Movement Disorder Society-Unified Parkinson's Disease Rating Scale (MDS-UPDRS) motor items. Results: Groups were age-homogeneous (P>.05). Participants with PD had significantly higher scores on MDS-UPDRS Part III in the OFF state compared to the ON state of medication (mean difference 2.86, 95% CI 2-12; P<.05). RT was generally shorter and AV higher in controls than in participants with PD. In the PD group, RT decreased and AV increased from OFF to ON states, reflecting an improvement in motor performance. Significant differences in RT were observed between groups in all phases of the game, with effect sizes ranging from small to moderate (η2=0.0239-0.0650). AV differed markedly between groups in phase 4, with a large effect size (η2=0.404). Correlation analyses revealed weak positive associations between RT and MDS-UPDRS items, while AV showed strong negative correlations with each motor item and the summary score for bradykinesia. Conclusions: This study proposes a new method to assess slowness in PD by using inertial sensor data to extract objective motor measures (RT and AV) during a serious game, allowing continuous and quantitative evaluation beyond traditional clinical scales and tests. The findings demonstrate that RT and AV extracted from gameplay can detect slowness-related motor differences, supporting the RehaBEElitation serious game as an alternative and objective digital biomarker, with potential applications in both clinical and home-based monitoring of symptom progression.