Advancing Brain State Analysis: A Phase-based Energy Landscape Method
Poster No:
1425
Submission Type:
Abstract Submission
Authors:
Ziyan Deng1, Jiahui Shi1, Chen Ran1, Ting Ma2, Chenfei Ye2
Institutions:
1Department of Electronic and Information Engineering, Harbin Institute of Technology (Shenzhen), ShenZhen, GuangDong, 2School of Biomedical Engineering, Harbin Institute of Technology (Shenzhen), ShenZhen, GuangDong
First Author:
Ziyan Deng
Department of Electronic and Information Engineering, Harbin Institute of Technology (Shenzhen)
ShenZhen, GuangDong
Department of Electronic and Information Engineering, Harbin Institute of Technology (Shenzhen)
ShenZhen, GuangDong
Co-Author(s):
Jiahui Shi
Department of Electronic and Information Engineering, Harbin Institute of Technology (Shenzhen)
ShenZhen, GuangDong
Department of Electronic and Information Engineering, Harbin Institute of Technology (Shenzhen)
ShenZhen, GuangDong
Chen Ran
Department of Electronic and Information Engineering, Harbin Institute of Technology (Shenzhen)
ShenZhen, GuangDong
Department of Electronic and Information Engineering, Harbin Institute of Technology (Shenzhen)
ShenZhen, GuangDong
Ting Ma
School of Biomedical Engineering, Harbin Institute of Technology (Shenzhen)
ShenZhen, GuangDong
School of Biomedical Engineering, Harbin Institute of Technology (Shenzhen)
ShenZhen, GuangDong
Chenfei Ye
School of Biomedical Engineering, Harbin Institute of Technology (Shenzhen)
ShenZhen, GuangDong
School of Biomedical Engineering, Harbin Institute of Technology (Shenzhen)
ShenZhen, GuangDong
Introduction:
Recent progress in the analysis of brain dynamics, particularly through the exploration of brain states (Greene et al., 2023), has been substantial. As research into brain metastability expands, new dynamic analysis methods, such as the LEiDA method (Cabral et al., 2017; Deco et al., 2019), have emerged. Although LEiDA has made valuable contributions in the area of translational neuroscience, it is not without limitations, including the incapacity to measure potential energy-a critical determinant in the dynamics of state transitions within the self-organization processes at rest. Drawing from the energy landscape theory (Ezaki et al., 2017), we posit that the intrinsic brain states should achieve equilibrium between metastable characterization and the intrinsic dynamics of state transitions. Therefore, we introduced a computational approach, named Energy Metastable State Analysis (EMSA), to identify the brain states that better represent the coordination patterns within the metastable phases of neural dynamics.
Methods:
Resting-state and task-based fMRI data from 590 healthy adult participants in the S1200 Human Connectome Project (HCP) were obtained. Participants performed seven cognitive tasks (Motor, Emotion, Language, Relational, Gambling, Social, Wm) and rest periods, each with two runs, differing in phase encoding direction (LR or RL). We performed brain state identification on fMRI data using EMSA (See Fig. 1 for methodological details), and assessed its performance against LEiDA and Energy Landscape Analysis (ELA) based on criteria including task differentiation, test-retest reliability, and task prediction capabilities. Specifically, we compared KL divergence of the same subject in the probability state space across different tasks to evaluate which method provides the highest task differentiation. Discriminability (Bridgeford et al., 2020) measures the overall consistency and differentiability of observations by computing the distance between all pairs of subjects and calculating the fraction of time (MNR) that a within-subject distance is smaller than between-subject distance. Prediction performance was assessed using a random forest model with accuracy and validated through ten-fold cross-validation.
Results:
(1) KL divergence varied across methods and subjects. We calculate the KL divergence for every two pairs of task states. The results show that task differentiation of LEiDA method improved with cluster number k (Fig. 2b). Considering four clustering indicators, LEiDA method identified k=7 as optimal. When choosing k=7, EMSA showed higher KL divergence compared to LEiDA and ELA in all 28 comparison cases (p < 0.001, Fig. 2a). (2) We used occurrence frequency as the feature to predict eight states (rest and seven tasks), and found that EMSA outperforms others on prediction accuracy (Fig. 2d). Furthermore, our analysis revealed that EMSA achieved the highest accuracy in predicting resting states, suggesting its superiority over alternative methods in discerning resting states from other tasks. This finding was further supported by pairwise classification analyses comparing rest with various tasks (Fig. 2e). (3) We calculated discriminability using the analysis results of two similar scans (LR-RL), demonstrating EMSA's remarkable advantages in test-retest reliability (Fig. 2f). In addition, we found a significant positive correlation between time series length and test-retest reliability (Fig. 2g). Notably, language tasks deviated from the trend, possibly due to their high pattern complexity (Capouskova et al., 2022).
Conclusions:
Our study proposes a novel approach to represent brain states based on fMRI scans, surpassing existing methods in terms of task differentiation, discriminability, and predictive performance. Future research should delve into the translational potential of this approach in various demographic groups, and endeavor to uncover related biomarkers for the diagnosis and monitoring of a range of neurological disorders.
Modeling and Analysis Methods:
fMRI Connectivity and Network Modeling 1
Methods Development 2
Task-Independent and Resting-State Analysis
Keywords:
FUNCTIONAL MRI
Meta- Analysis
Other - brain dynamics; state identification; metastable state；energy landscape；task prediction
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?
Provide references using APA citation style.
2. Cabral, J. (2017). Cognitive performance in healthy older adults relates to spontaneous switching between states of functional connectivity during rest. Scientific reports, 7(1), 5135.
3. Capouskova, K. (2022). Modes of cognition: Evidence from metastable brain dynamics. NeuroImage, 260, 119489.
4. Deco, G. (2019). Awakening: Predicting external stimulation to force transitions between different brain states. Proceedings of the National Academy of Sciences, 116(36), 18088-18097.
5. Ezaki, T. (2017). Energy landscape analysis of neuroimaging data. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 375(2096), 20160287.
6. Greene, A. S. (2023). Why is everyone talking about brain state?. Trends in Neurosciences, 46(7), 508-524.