High Variability in Node Association Seen in Community Analysis of the Programmers’ Connectome
Poster No:
1780
Submission Type:
Abstract Submission
Authors:
Ronnie Krupnik1, Yaniv Assaf2
Institutions:
1Tel Aviv University, Giv'atayim, Outside the U.S. & Canada, 2Tel Aviv University, Tel Aviv, Outside the U.S. & Canada
First Author:
Co-Author:
Introduction:
Programming is a complex cognitive skill that engages multiple brain modules, including language, logic, arithmetics and more. Writing or examining code entails activation in networks related to attention, working memory and planning, mostly in the right hemisphere, as well as language related areas (Krueger et al, 2020; Peitek et al, 2018). As such, programmers are expected to differ from the general population not only in brain structure but also in the connections between the regions supporting these skills. In this study we investigate the differences in network structure between expert programmers and naive subjects. We suggest a community-based analysis, examining the variation in community structure, stability and the degree to which group-based partitions represent individual brain connectivity patterns.
Methods:
We scanned 172 healthy subjects , 73 of which were programming professionals (experts), while the rest had no experience with code (naive). Structural and diffusion-weighted MRI scans were obtained, and multi-tissue multi-shell diffusion tractography was performed using MrTrix software (Tournier et al., 2004 & 2019). Connectivity matrices were derived based on the Brainnetome Atlas (Fan et al., 2016).
Each subject's connectome was parcellated into communities using the Clauset-Newman-Moore greedy modularity maximization algorithm. For each group we created group-wise partitionings by (a) the subject's partition with the highest mutual information (MI) to others in the group, and (b) the mean connectivity matrix of each group.
We analyzed which nodes are assigned different communities between the groups. Statistical significance was determined by the frequency with which nodes were reassigned across random groups in a permutation test. We then computed a within-group stability value for each node based on (a) the proportion of times each node was assigned to the most common community and (b) the entropy of the assignment vector for each node within the group.
Last, we applied the expert group partitioning to each individual's network, and compared modularity measures, such as within-community density and efficiency, using t-tests between the groups.
Each subject's connectome was parcellated into communities using the Clauset-Newman-Moore greedy modularity maximization algorithm. For each group we created group-wise partitionings by (a) the subject's partition with the highest mutual information (MI) to others in the group, and (b) the mean connectivity matrix of each group.
We analyzed which nodes are assigned different communities between the groups. Statistical significance was determined by the frequency with which nodes were reassigned across random groups in a permutation test. We then computed a within-group stability value for each node based on (a) the proportion of times each node was assigned to the most common community and (b) the entropy of the assignment vector for each node within the group.
Last, we applied the expert group partitioning to each individual's network, and compared modularity measures, such as within-community density and efficiency, using t-tests between the groups.
Results:
While overall the group-wise partitions seemed similar, notable differences emerged. In the mean matrix grouping, the naive group exhibited an additional bilateral community, where the medial-frontal network included nodes from both frontal lobes, which were assigned to hemispheric networks in the expert group.
Some temporal nodes were assigned to frontal communities in the expert group but to temporal-occipital ones in the naive group. Permutation testing further revealed that the expert group's partitioning displayed higher variability, making it a less reliable indicator of group membership.
To further understand this finding, we examined the difference of stability of each node between the groups. Significant differences were observed in areas including the right parahippocampal gyrus, superior temporal and insular areas and the precuneus (fig. 1).
When applying the programmers' partitioning to all subjects' connectome we found significant differences in within-community density between groups in the cerebellum and left-frontal communities (fig. 2).
Some temporal nodes were assigned to frontal communities in the expert group but to temporal-occipital ones in the naive group. Permutation testing further revealed that the expert group's partitioning displayed higher variability, making it a less reliable indicator of group membership.
To further understand this finding, we examined the difference of stability of each node between the groups. Significant differences were observed in areas including the right parahippocampal gyrus, superior temporal and insular areas and the precuneus (fig. 1).
When applying the programmers' partitioning to all subjects' connectome we found significant differences in within-community density between groups in the cerebellum and left-frontal communities (fig. 2).
·figure 1
·figure 2
Conclusions:
Programming is a complex skill, and as such involves not only specialized brain regions but also distinct network architecture. Our findings suggest that the variation in community structure and node allocation in the brain may reflect the neural bases of coding expertise, as programmers show a more lateralized network structure and a variation in association between left frontal and temporal nodes. The higher variability in the expert group's brain connectivity may indicate a more flexible and interconnected network, possibly reflecting a stronger or more complex hub architecture. This variation, rather than a simple structural difference, may be key to understanding how expert programmers' brains differ from naive individuals.
Higher Cognitive Functions:
Higher Cognitive Functions Other
Learning and Memory:
Skill Learning
Modeling and Analysis Methods:
Diffusion MRI Modeling and Analysis 2
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
White Matter Anatomy, Fiber Pathways and Connectivity 1
Novel Imaging Acquisition Methods:
Diffusion MRI
Keywords:
Cognition
Computational Neuroscience
Data analysis
MRI
Other - Tractography; connectome; graph theory; community analysis
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.
Peitek, N., et al. (2020), A Look into Programmers’ Heads. IEEE Transactions on Software Engineering, 46(4), 442-462.
Fan, L. et al. (2016), The Human Brainnetome Atlas: A New Brain Atlas Based on Connectional Architecture. Cerebral Cortex 26, 3508–3526 .
Tournier, J.-D.; Calamante, F., Gadian, D.G. & Connelly, A. (2004). Direct estimation of the fiber orientation density function from diffusion-weighted MRI data using spherical deconvolution. NeuroImage,, 23, 1176-1185
Tournier, J.-D.; Smith, R. E.; Raffelt, D.; Tabbara, R.; Dhollander, T.; Pietsch, M.; Christiaens, D.; Jeurissen, B.; Yeh, C.-H. & Connelly, A. (2019) MRtrix3: A fast, flexible and open software framework for medical image processing and visualisation. NeuroImage, 202, 116137