Résumé
Objective: In movement science, there are no recommendations regarding the number of trials (NoT) required to stabilize the intertrial variability of the cerebral hemodynamic response (CHR) in fNIRS block design. This study aimed to determine the minimum NoT required to stabilize CHR variability, based on Δ[HbO] and Δ[HbR] measured by fNIRS during a motor and a proprioceptive task.Methods: Twenty-six young, healthy, right-handed individuals (24±5 years, 13 females) performed a volitional opening/closing right-hand task (1-Hz metronome) and experienced vibration-induced illusory (80 Hz) arm flexion. Trials lasted 10-second with a [12-20]-second intertrial interval, and were presented 20 times. The order was counterbalanced. Continuous-wave fNIRS measurements (25 Hz) were localized on bi-hemispheric sensorimotor and frontoparietal areas. Bad channels were pruned (visually/QT-NIRS). Data were corrected (NIRSTORM) for movements, frequencies and superficial noise. Based on cumulative moving average over trials (Δ[HbO], Δ[HbR), task ([5-15] seconds) standard deviations were extracted. The response was estimated with a Generalized Additive Model. Finally, splines (<0.01) detected the plateau, resulting in the minimum NoT.Results: In the volitional condition, minimum NoT for the contralateral sensorimotor area was 9/13 (HbO/HbR). In the illusory condition, the minimal NoT for the contralateral sensorimotor area was 5/5, and 9/10 for the frontoparietal areas. Conclusion: The results indicate that the signal appears to stabilize at a different NoT for voluntary and illusory movements, depending on the areas and chromophores. This finding should be considered in fNIRS studies to ensure detecting CHR, and limiting fatigability/habituation. This paves the way for future standardization of fNIRS movement science block design studies.