Frontal-parietal Neural Similarity Between Tasks Predicts Individual-level Training Transfer
Poster No:
832
Submission Type:
Abstract Submission
Authors:
Xiyue Chen1, Siyuan Lyu2, Junjiao Feng3, Ying Cai1
Institutions:
1Department of Psychology and Behavioral Sciences, Zhejiang University, Hangzhou, Zhejiang, 2Zhejiang University, Jiaxing, Zhejiang, 3Faculty of Psychology, Tianjin Normal University, Tianjin, Tianjin
First Author:
Co-Author(s):
Introduction:
Training transfer indicates that training on one task can improve other untrained tasks. Classical transfer theories suggest that a greater similarity between tasks indicates a higher likelihood of transfer. However, these theories cannot explain the individual differences in training transfer. This study explores the underlying mechanism of individual differences in training transfer through a novel perspective of neural pattern similarity (Cohen et al., 2014; Ito et al., 2022).
Methods:
Fifty subjects participated in our study (24 females, age = 18.3 ± 1.7 years). First, participants completed n-back and Stop-signal tasks (SST) while undergoing fMRI scanning. In the n-back task, participants made judgments about whether the current letter matched the one presented n-trials ago. In the SST task, participants pressed different buttons according to the directions of arrows. However, in 25% of trials, participants needed to inhibit their responses when the arrow turned red after the onset of presentations. Then, participants performed difficulty-adaptive n-back training (5 days, 30 minutes per day) and completed SST tasks before and after training sessions.
For the n-back task, we used IE = RT/accuracy as the behavioral index (Liu et al., 2017). For the SST task, we used SSRT = quantile RT – mean SSD (quantile RT: ascending the RT in go trials and choosing the RT at the inhibition rate percentile; stop signal delay (SSD): the durations between the onsets of go trials and the stop trials) as the behavioral index (Cai et al., 2016); and the percentage change of SSRT served as the transfer index. To calculate neural similarity across tasks, we applied a general linear model (GLM) to obtain the brain activities during the n-back (2-back vs. 0-back) and SST tasks (stop vs. go), respectively. Next, we used the Power 264 template and did Spearman correlations to estimate the neural similarity for each subnetwork while participants performed the two tasks. To improve the signal-to-noise ratio, we focused on the voxels that were activated in at least one task (z > 2.3). Furthermore, we divided participants into two groups (n = 25) depending on their neural similarity in each subnetwork and compared the transfer indexes between groups; Meanwhile, we also performed Pearson correlations to examine whether neural similarity predicted the transfer effect.
For the n-back task, we used IE = RT/accuracy as the behavioral index (Liu et al., 2017). For the SST task, we used SSRT = quantile RT – mean SSD (quantile RT: ascending the RT in go trials and choosing the RT at the inhibition rate percentile; stop signal delay (SSD): the durations between the onsets of go trials and the stop trials) as the behavioral index (Cai et al., 2016); and the percentage change of SSRT served as the transfer index. To calculate neural similarity across tasks, we applied a general linear model (GLM) to obtain the brain activities during the n-back (2-back vs. 0-back) and SST tasks (stop vs. go), respectively. Next, we used the Power 264 template and did Spearman correlations to estimate the neural similarity for each subnetwork while participants performed the two tasks. To improve the signal-to-noise ratio, we focused on the voxels that were activated in at least one task (z > 2.3). Furthermore, we divided participants into two groups (n = 25) depending on their neural similarity in each subnetwork and compared the transfer indexes between groups; Meanwhile, we also performed Pearson correlations to examine whether neural similarity predicted the transfer effect.
Results:
Although behavioral performance on the n-back and SST tasks revealed a positive correlation in pre-training condition (r = 0.43, p < 0.01), SST performance did not improve after n-back training (t = -0.74, p = 0.46; Figure 1B). These results indicated that behavioral correlations between tasks cannot predict training transfer. However, in the frontoparietal network (FPN), the high neural similarity group exhibited significantly greater transfer effects compared to the low neural similarity group (t = 2.14, p = 0.03). Moreover, among participants who demonstrated transfer effects (i.e. SST is smaller in the post-training test, n = 30), neural similarity within the FPN positively predicted the transfer effects (r = 0.291, p = 0.04; Figure 1C). No such correlations were observed in other subnetworks (rs <0.125, ps > 0.387). These findings suggest that similarity in neural activity patterns within the FPN across tasks specifically predicts individual training transfer.
Conclusions:
This study first demonstrates that neural similarity within the FPN could predict training transfer at the individual level, suggesting that the similarity of higher-order cognitive control networks across tasks determines the training transfer. These findings deepen our understanding of training transfer and provide evidence for establishing personalized training strategies.
Learning and Memory:
Working Memory
Learning and Memory Other 1
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI)
Novel Imaging Acquisition Methods:
BOLD fMRI 2
Keywords:
FUNCTIONAL MRI
Memory
Other - training transfer, neural similarity
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
Cohen, M. A., Konkle, T., Rhee, J. Y., Nakayama, K., & Alvarez, G. A. (2014). Processing multiple visual objects is limited by overlap in neural channels. Proceedings of the National Academy of Sciences, 111(24), 8955-8960.
Ito, T., Yang, G. R., Laurent, P., Schultz, D. H., & Cole, M. W. (2022). Constructing neural network models from brain data reveals representational transformations linked to adaptive behavior. Nature Communications, 13(1), 1–16.
Liu, J., Xia, M., Dai, Z., Wang, X., Liao, X., Bi, Y., & He, Y. (2016). Intrinsic Brain Hub Connectivity Underlies Individual Differences in Spatial Working Memory. Cerebral Cortex, 27, 5496–5508.