Oculomotor dance learning task: Implications for audio-visual cued spatial learning

Learning dance of a motor sequence involves the coordination of both oculomotor and manual motor systems through the practiced repetition of a fixed sequence of actions, resulting in automatized execution of movement through habit learning. This study aims to address whether a sequence-based learning paradigm centered on the visual-motor system can feasibly be measured while listening to music (Bar and DeSouza 2016). It aims to develop a new visual-motor-based learning paradigm with music, potentially promoting neuroplasticity and creating new interventional tools, building upon prior research that shows behavioural and putative neural changes following dance-based neurorehabilitation in people with Parkinson's disease (Bearss et al. 2024). Eye movements of 10 participants (8 female, 2 male) were tracked using the Eyelink 1000 Plus system during a 68-second eye-dance sequence. The experiment consisted of a learning phase, where participants observed the sequence five times with 30-second breaks, and a performance phase, where they performed the sequence five times from memory on a grey screen without visual cues. Music was incorporated into both phases to aid memorization of the 4 spatial locations. After each performance, the participant was shown a visual reinforcer and asked for their thoughts on how well they executed the dance. A visual reinforcer flashes one of three different colours: red, yellow, or green. Each colour corresponds to how many steps in the dance a participant performed correctly, with key points being: under one third, between one to two thirds, and over two thirds of total steps correct. Participants were scored based on timing of the steps as well for exact (1.00), good (0.66), slightly off (0.33) or missed (0) steps. Data was analyzed using R4.3.1, MATLAB, and Experiment Builder: Data Viewer software. Results showed a significant improvement in performance accuracy between the first session (g1; M = 40%, SD = 7.2%) and the last session (g5; M = 69.7%, SD = 22.8%). A repeated-measures ANOVA revealed a significant main effect of session on performance accuracy, F(4, 36) = 6.99, p < 0.001, 2G = 0.26, indicating that accuracy significantly improved over sessions. Post-hoc Bonferroni comparisons showed that accuracy in later sessions was significantly higher than earlier sessions, suggesting a defined learning curve and consolidation of performance pattern across repeated practice. Similarly, there was significant improvement in timing accuracy between the first session g1; M = 0.29, SD = 0.06) and the fifth session (g5; M = 0.46, SD = 0.12). A repeated-measures ANOVA revealed a significant main effect of session on timing precision, F(4, 36) = 11.67, p < 0.001, 2G = 0.25, indicating significant improvements in temporal control and coordination over sessions. Post-hoc Bonferroni comparisons showed that timing precision significantly improved between early and late sessions (e.g, g1-g4, p <0.01; g1-g5, p < 0.001), suggesting a defined learning curve and increase in precision across repeated practice. These findings suggest that visual-motor-based interventions have the potential to enhance motor and non-motor symptoms like depression and anxiety for neurodegenerative diseases such as Parkinson's Disorder (PD). The results provide a foundation for developing targeted therapies that integrate learning paradigms to improve functional outcomes, warranting further exploration of their long-term efficacy.