Poster D96, Monday, March 27, 8:00 – 10:00 am, Pacific Concourse
Post-practice resting-state functional connectivity predicts the benefit of contextual interference on motor learning
Chien-Ho Lin1, Ho-Ching Yang1, Barbara Knowlton2, Shin-Leh Huang1, Ming-Chang Chiang1; 11National Yang-Ming University, Taiwan, 2UCLA
Interleaved practice (IP) of motor sequences generally leads to better learning than repetitive practice (RP). Moreover, IP was found to be associated with greater fMRI blood-oxygen-level dependent (BOLD) signals in the sensorimotor areas during practice, but lower BOLD signals during retention. Here we investigated whether post-practice resting-state functional connectivity predicts the learning benefit of IP. 26 healthy adults (11M/15F, age = 23.3±1.3 years) practiced two sets of three sequences arranged in a Repetitive or an Interleaved order over 2 days, followed by a retention test on Day 5 to evaluate learning. On each practice day, fMRI data were acquired in a resting state after practice. The resting-state fMRI data was decomposed using a group-level spatial independent component analysis (ICA), yielding 9 independent components (IC) matched to the ventral and dorsal default-mode networks (DFN), primary visual (2 ICs), sensorimotor, left and right executive-control, posterior salience, and the language networks. The functional connectivity strength between a voxel and an IC was gauged by a z-score. Greater RP-minus-IP difference in connectivity strength was found on Day 1 in the left posterior inferior temporal gyrus belonging to the ventral DFN, while the IP-minus-RP difference in connectivity strength in the left premotor cortex (Brodmann area 6; the sensorimotor network) on Day 1 was positively correlated with the RP-minus-IP difference in the response time on Day 5. Our findings suggest that resting brain networks in the early phase of consolidation may be a biomarker of enhanced learning following interleaved practice.
Topic Area: LONG-TERM MEMORY: Skill learning