Brain fingerprinting via EC eigenmodes: unveiling unimodal networks’ role in near-critical dynamics
Poster No:
1389
Submission Type:
Abstract Submission
Authors:
Giorgia Baron1, Claudia Tarricone1,2, Danilo Benozzo1, Giacomo Baggio1, Sandro Zampieri1, Alessandro Chiuso1, Alessandra Bertoldo1,2
Institutions:
1Department of Information Engineering, University of Padua, Padua, Italy/Padua, 2Padova Neuroscience Center, University of Padua, Padua, Italy, Italy
First Author:
Co-Author(s):
Claudia Tarricone
Department of Information Engineering, University of Padua|Padova Neuroscience Center, University of Padua
Padua, Italy/Padua|Padua, Italy, Italy
Department of Information Engineering, University of Padua|Padova Neuroscience Center, University of Padua
Padua, Italy/Padua|Padua, Italy, Italy
Alessandra Bertoldo
Department of Information Engineering, University of Padua|Padova Neuroscience Center, University of Padua
Padua, Italy/Padua|Padua, Italy, Italy
Department of Information Engineering, University of Padua|Padova Neuroscience Center, University of Padua
Padua, Italy/Padua|Padua, Italy, Italy
Introduction:
The "fingerprinting" of individuals based on their brain connectome is a key goal in neuroscience. Recent studies showed higher identification accuracy from default mode (DMN) and executive control (CONT) networks (Griffa et al., 2022). However, these methods are limited by their use of undirected functional connectivity (FC), not capturing connections' asymmetries. To address this, asymmetric effective connectivity (EC) offers a more comprehensive representation of brain interactions. Additionally, EC large-scale eigendecomposition, involving complex eigenmodes, could provide a richer framework to characterize human fingerprinting by capturing dynamic flow directionality across multiple spatiotemporal scales (Chopra et al., 2023).
Methods:
Two rsfMRI runs of HCP Healthy Young Adult dataset (144 subjects) (Van Essen et al., 2012) underwent preprocessing and parcellation steps with the clustered Schaefer functional atlas (62 ROIs/7 RSNs), plus 12 subcortical and cerebellar regions (AAL2). For each subject, the first run was used as test session and the second one as retest. Each individual EC matrix obtained via sparse DCM (Prando et al., 2020) was eigendecomposed as ∑(i=1...n)=λi*vi*wi_T: each ith complex eigenmode (i.e. right vi and left wi_T eigenvectors associated to the eigenvalue λi of the linear system) represents low-dimensional axes for neural trajectories evolution at various temporal scales. Each ith eigenmode's dynamic flow is associated with a specific amount of kinetic energy linked to distinct levels of dissipative energy (i.e. Ei_Re=Re(λi)^2/(-2Re(λi)) and solenoidal energy (i.e. Ei_Im=Im(λi)^2/(-2Re(λi)) (Friston et al., 2021). After removing the λi with Re(λi)<−0.5 (associated to noise-related dynamics), we partitioned the complex eigenspace based on the 'median-split' of the flow kinetic energy (Preti et al., 2019). Specifically, the area under the curve (AUC) of Ei_Re and Ei_Im was calculated, and cutoff thresholds were identified along Re(λi) and Im(λi) to split the AUC into two comparable parts. This resulted into three eigenmodes' energetic ranges (Fig. 1a): dissipative R1 (Ei_Re>Ei_Im); intermediate R2 (Ei_Re≈Ei_Im); solenoidal R3 (Ei_Re<Ei_Im). Then, to statistically characterize the EC matrices' causal structure, we used the closed-form decomposition outlined in (Benozzo et al., 2024), yielding S (i.e. skew-symmetric differential cross-covariance matrix), encoding directional asymmetries, and Σ (i.e. symmetric covariance matrix). We assessed the fingerprinting power of S and Σ, projected onto R1, R2, and R3 eigenmodes, computing differential identifiability (Idiff), intraclass correlation coefficient (ICC) and its column-wise strength (ICCstrength), alongside the success rate (SR) of subject identification (Amico et al., 2018). Statistical differences were evaluated via ANOVA and post-hoc tests.
Results:
The results revealed an increasing complementary capacity of S and Σ fingerprinting across energetic regimes. Coherent Idiff and SR patterns across ranges further confirmed Σ robust and stable fingerprinting in both R2 and R3, even in different areas, whereas S showed statistically significant differences in fingerprinting power across energetic transitions, peaking in R3 (Fig. 1b). Σ exhibited higher ICCstrength in DMN and CONT regions in R2, as well as in some unimodal regions (particularly sensory motor) in R3. S-ICCstrength demonstrated increasing values in unimodal areas and the thalamus, complementing those of Σ, when transitioning from R2 to R3 (Fig. 2).
Conclusions:
By leveraging the decomposition of asymmetric EC, inherently tied to directional flow via its complex eigenvalues, we revealed the high fingerprinting capacity of unimodal networks (Sareen et al., 2021), a contribution often overlooked in traditional fMRI analyses focusing on static FC or structural harmonics constrained to real eigenvalues. These results highlight the importance of near-critical, solenoidal dynamics in shaping individual brain signatures.
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural) 2
fMRI Connectivity and Network Modeling 1
Task-Independent and Resting-State Analysis
Novel Imaging Acquisition Methods:
BOLD fMRI
Keywords:
Computational Neuroscience
FUNCTIONAL MRI
Other - Dynamic Causal Modelling; Effective connectivity; Complex eigenmodes; Kinetic energy; Near-critical dynamics; Brain fingerprinting; Unimodal networks
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. Benozzo, D., Baggio, G., Baron, G., Chiuso, A., Zampieri, S., & Bertoldo, A. (2024). Analyzing asymmetry in brain hierarchies with a linear state-space model of resting-state fMRI data. Network Neuroscience, 8(3), 965–988.
3. Chopra, S., Zhang, X.-H., & Holmes, A. J. (2023). Wave-like properties of functional dynamics across the cortical sheet. Neuron, 111(8), 1171–1173.
4. Friston, K. J., Fagerholm, E. D., Zarghami, T. S., Parr, T., Hipólito, I., Magrou, L., & Razi, A. (2021). Parcels and particles: Markov blankets in the brain. Network Neuroscience, 5(1), 211–251.
5. Griffa, A., Amico, E., Liégeois, R., Van De Ville, D., & Preti, M. G. (2022). Brain structure-function coupling provides signatures for task decoding and individual fingerprinting. NeuroImage, 250, 118970.
6. Liégeois, R., Santos, A., Matta, V., Van De Ville, D., & Sayed, A. H. (2020). Revisiting correlation-based functional connectivity and its relationship with structural connectivity. Network Neuroscience, 4(4), 1235–1251.
7. Prando, G., Zorzi, M., Bertoldo, A., Corbetta, M., Zorzi, M., & Chiuso, A. (2020). Sparse DCM for whole-brain effective connectivity from resting-state fMRI data. NeuroImage, 208, 116367.
8. Preti, M. G., & Van De Ville, D. (2019). Decoupling of brain function from structure reveals regional behavioral specialization in humans. Nature Communications, 10(1), 4747.
9. Sareen, E., Zahar, S., Ville, D. V. D., Gupta, A., Griffa, A., & Amico, E. (2021). Exploring MEG brain fingerprints: Evaluation, pitfalls, and interpretations. NeuroImage, 240, 118331.
10. Van Essen, D. C., Ugurbil, K., Auerbach, E., Barch, D., Behrens, T. E. J., Bucholz, R., Chang, A., Chen, L., Corbetta, M., Curtiss, S. W., Della Penna, S., Feinberg, D., Glasser, M. F., Harel, N., Heath, A. C., Larson-Prior, L., Marcus, D., Michalareas, G., Moeller, S., … Yacoub, E. (2012). The Human Connectome Project: A data acquisition perspective. NeuroImage, 62(4), 2222–2231