A novel method for synchronized rhythmic production of finger movements and the effects of auditory stimuli and feedback

Manual dexterity relies on accurate and timely coordinated finger movements. Rhythmic finger-movement training can support motor practice, but solutions that combine precise timing with synchronized speech or music feedback remain scarce. We developed Rhythm Karaoke, a rhythmic finger-tapping platform that synchronizes auditory stimuli with finger taps recorded by a manual dexterity device.

The platform provides adaptive sound synthesis, multiple stimulus types (notes vs. speech; melodic vs. monotone), real-time auditory feedback, and tapping-error quantification. To validate the method, we tested whether melodic and semantic content (speech) improve rhythmic precision relative to monotone non-semantic notes, and whether auditory feedback alters performance, motivation, and preference. Healthy participants (n = 30) reproduced rhythmic sequences from three French songs in a within-subject design crossing Melody (melodic / monotone), Semantic (speech / notes), and Feedback (present / absent).

Tapping error was defined as the average absolute difference between expected and observed proportional inter-tap intervals, expressed as a percentage of song duration. Subjective experience was assessed with the Intrinsic Motivation Inventory. Test-retest reliability was evaluated in a separate group of healthy participants (N = 20) for the easiest song in the melodic-note and melodic-speech conditions.

Tapping error differed significantly across songs. ANOVA showed significant effects of Melody and Semantic content and a Melody × Semantic interaction: melody improved performance only for note sequences, and a significant Song × Melody × Semantic interaction showed that the size of the melodic benefit depended on the song. Auditory feedback showed a significant Semantic × Feedback interaction: it tended to reduce tapping error for notes (non-significant after Bonferroni correction) but significantly increased error for speech.

Participants preferred melodic conditions, reporting lower perceived tension and higher interest and satisfaction. Reliability was higher for melodic-note than melodic-speech conditions, with excellent test-retest reliability for melodic-note with auditory feedback. These results indicate that melody and semantic content can improve rhythmic precision and engagement, and they support the potential of Rhythm Karaoke for manual dexterity training and neurorehabilitation.

We thank Christophe d’Alessandro and the Lutheries–Acoustique–Musique (LAM) team at Sorbonne Université for their valuable discussions and technical support in the development of the Rhythm Karaoke platform. We also thank all study participants for their time and commitment. Institute of Psychiatry and Neuroscience of Paris, Université Paris Cité, INSERM U1266, Paris, F-75014, France De Vinci Research Center, De Vinci Higher Education, Paris, France P.G.L. is a founding member of the start-up company Dextrain and holds a patent related to manual dexterity measurement (WO2016184935A3).