Brain Functional Topologies under Naturistic Stimuli using Persistent Homology
Poster No:
1203
Submission Type:
Abstract Submission
Authors:
Abdul Rehman Khan1, Muhamad Shahzaib2, Sadia Shakil1, Raymond Tong1
Institutions:
1Department of Biomedical Engineering, The Chinese University of Hong Kong, Sha Tin, Hong Kong, 2Biosignal Processing and Computational Neuroscience Lab, Institute of Space Technology, Islamabad, Pakistan
First Author:
Abdul Rehman Khan, Mr
Department of Biomedical Engineering, The Chinese University of Hong Kong
Sha Tin, Hong Kong
Department of Biomedical Engineering, The Chinese University of Hong Kong
Sha Tin, Hong Kong
Co-Author(s):
Muhamad Shahzaib
Biosignal Processing and Computational Neuroscience Lab, Institute of Space Technology
Islamabad, Pakistan
Biosignal Processing and Computational Neuroscience Lab, Institute of Space Technology
Islamabad, Pakistan
Sadia Shakil
Department of Biomedical Engineering, The Chinese University of Hong Kong
Sha Tin, Hong Kong
Department of Biomedical Engineering, The Chinese University of Hong Kong
Sha Tin, Hong Kong
Raymond Tong
Department of Biomedical Engineering, The Chinese University of Hong Kong
Sha Tin, Hong Kong
Department of Biomedical Engineering, The Chinese University of Hong Kong
Sha Tin, Hong Kong
Introduction:
Understanding brain function, under naturalistic stimuli (e.g. video or audio), has recently gained popularity (Eickhoff,2020). The fMRI volumes are used to identify brain functional networks through set of voxels co-activated during stimulus (Gore,2003). Brain network structure alters with stimuli, resulting in their spatiotemporal dynamics. Recently, topological data analysis (TDA) has shown promising results for studying brain topologies (structures) (Wang,2024; Catanzaro,2024) using Persistent Homology (PH) (Munch,2017), which encodes point cloud structure in 2D Betti curve. By comparing these diagrams across brain networks, the structural similarity is quantified. Our research explores the spatio-temporal dynamics of brain networks by answering following questions: (1) Do the brain network topologies repeat across time, and (2) Do the brain network topologies repeat across subjects? We employed PH to extract the topological structures from DMN and visual networks and understand the evolution of brain activity across subjects and sessions.
Methods:
Aligned with our research objectives we used fMRI-video dataset (Wen,2018), collected in multi-subject and multi-session experiment. Three subjects watched 18 video segments twice. We used pre-processed T2-weighted fMRI dataset, removed resting state volumes, and worked with 240 fMRI volumes for each segment. We studied the spatiotemporal dynamics of brain activity in the DMN and visual networks. For spatial and temporal dynamics, persistent homology was applied along volume and voxel axis, respectively (figure 1). The fMRI signal, consisting of volumes was flattened to obtain a matrix with shape N×M, where N=#volumes and M=#voxels. For temporal (spatial) analysis, voxel (volume) wise correlation matrix was computed and correlated distance (CD) was defined as (1-correlation)/2. The CD effectively captures the notion of distance by mapping [-1, +1] correlation range to [0, 1] distance range, thus smaller distance indicates higher positive correlation and vice versa. The persistent homology (PH) was applied on voxels (volumes) point cloud using CD as a distance metric. From PH output, the Betti curves were constructed and AUC under Betti curve was obtained. We applied our persistent homology procedure separately per segment, across all subjects, views, and both networks.
·Procedure to compute spatial and temporal dynamics of fMRI data using Persistent Homology and Betti Curves.
Results:
The inter-subject and inter-session spatio-temporal topologies in brain network are quantified using Betti curves and AUC distribution in figure 2. The Betti curves were averaged across segments for each unique (subject, view, network) tuple, while AUC was computed for each Betti curve. The spatial (temporal) dynamics are compared in top (bottom) subplots, while each column corresponds to a subject. Different colors and line styles provide network and session contrast in plots. The Betti curves encode the spatial (temporal) structure of brain by recording the merging radii in volumes (voxels) point cloud. A network with consistent topologies will merge rapidly giving Betti curves with small AUC. The average Betti curves for the visual network are lower than DMN, indicating topological consistency in visual network. The distribution of AUC (without averaging) also provides similar evidence of topological consistency in visual network and DMN.
·Results of the proposed methodology, visual network has better topological consistency then DMN across spatial and temporal dynamics of brain topologies.
Conclusions:
With the help of persistent homology (PH), we tried to understand the spatiotemporal dynamics in DMN and visual networks under naturistic stimuli watching across multiple subjects and views. The spatial (temporal) dynamics were captured by applying persistent homology on volume (voxel) wise correlation distance metrices. The spatio-temporal dynamics of brain functional networks were compared by comparing average Betti curves and AUC distributions, in multi-view and multi-subject setting. For both research objectives, the results provide evidence of higher topological similarities in spatio-temporal dynamics for the visual network compared to DMN.
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI)
Connectivity (eg. functional, effective, structural) 1
fMRI Connectivity and Network Modeling
Methods Development 2
Other Methods
Keywords:
Computational Neuroscience
Computing
Data analysis
FUNCTIONAL MRI
Informatics
Structures
Other - spatio-temporal dynamics; topological data analysis; persistent homology; naturistic stimuli
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.
2. Eickhoff, S. B. (2020). Towards clinical applications of movie fMRI. NeuroImage.
3. Gore, J. C. (2003). Principles and practice of functional MRI of the human brain. The Journal of clinical investigation, 9.
4. Munch, E. (2017). A User’s Guide to Topological Data Analysis. Journal of Learning Analytics. Journal of Learning Analytics.
5. Wang, Z.-m. (2024). Emotion recognition based on phase-locking value brain functional network and topological data analysis. Neural Computing and Applications.
6. Wen, H. (2018). Neural encoding and decoding with deep learning for dynamic natural vision. Cerebral Cortex, 28.