Intermanual transfer of visuomotor adaptation is related to awareness

Abstract

Previous studies compared the effects of gradual and sudden adaptation on intermanual transfer to find out whether transfer depends on awareness of the perturbation. Results from different groups were contradictory. Since results of our own study suggest that awareness depends on perturbation size, we hypothesize that awareness-related intermanual transfer will only appear after adaptation to a large, sudden perturbation but not after adaptation to a small sudden perturbation or a gradual perturbation, large or small. To confirm this, four groups (S30, G30, S75, G75) of subjects performed out-and-back reaching movements with their right arm. In a baseline block, they received veridical visual feedback of hand position. In the subsequent adaptation block, feedback was rotated by 30 deg (S30, G30) or 75 deg (S75, G75). This rotation was either introduced suddenly (S30, S75) or gradually in steps of 3 deg (G30, G75). After the adaptation block, subjects did an awareness test comprising exclusion and inclusion conditions. The experiment concluded with an intermanual transfer block, in which movements were performed with the left arm under rotated feedback, and a washout block again under veridical feedback. We used a hierarchical Bayesian model to estimate individual movement directions and group averages. The movement directions in different conditions were then used to calculate group and individual indexes of adaptation, awareness, unawareness, transfer and washout. Both awareness and transfer were larger in S75 than in other groups, while unawareness and washout were smaller in S75 than in other groups. Furthermore, the size of awareness indices correlated to intermanual transfer across subjects, even when transfer was normalized to final adaptation level. Thus, we show for the first time that the amount of intermanual transfer directly relates to the extent of awareness of the learned perturbation.

Fig 1. Scheme of experimental apparatus.

Shown are tablet (T), screen (S) and mirror (M).

Fig 2. Mean angular movement errors of all blocks and groups.

Shown are baseline (BAS), adaptation (ADAP), exclusion (EX), refresh (R), inclusion (IN), intermanual transfer (TL) and de-adaptation (DA). Lines indicate across-subject means, and the shaded area display standard errors.

Fig 3. Adaptation.

Group means and their HDIs of posterior probability distributions for adaptation indices (A). Average difference from G30 (B) and S30 (C) of the last ten adaptation bins.

Fig 4. Awareness, unawareness, intermanual transfer and washout.

Group means and their HDIs of posterior probability distributions for awareness, unawareness, intermanual transfer and washout indices (A) as well as for normalized intermanual transfer (B).

Fig 5. Individual subjects performance for awareness, unawareness, intermanual transfer and washout.

Each line represents one subject and the range of the line is the HDI for that individual. Subjects are sorted by their mean score and coloured according to their group.

Fig 6. Correlations between awareness, unawareness, intermanual transfer and normalized intermanual transfer.

Correlations between the awareness and transfer (A) or normalized transfer (B) indices as well as correlations between the unawareness and transfer (C) and normalized transfer (D) indices for all participants.