Skip to main content
2024 BMES Annual Meeting Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.

Biomedical Imaging and Instrumentation

Biomedical Imaging and Instrumentation - Poster Session E

Poster G3 - Utilizing Entropy to Assess the Complexity of Brain Network Signals Acquired via fMRI

Saturday, October 26, 2024
10:00 AM - 11:00 AM EST
Location: Exhibit Hall E, F & G

Introduction:  The primary objective of this study is to examine how entropy is used in the analysis of functional magnetic resonance imaging (fMRI) signals.
The fMRI signal consists of sequential brain images taken over a specific period, using the intensity differences in blood-oxygen-level dependent (BOLD) signals, an indirect measure of neural activity that provides high spatial resolution [1].
Entropy is generally understood as a measure of unpredictability within a system. In time-based physiological signals, entropy quantifies the disorderliness in the time dimension [2]. According to the second law of thermodynamics, a system reaches maximum entropy at equilibrium [3]. However, living systems avoid this state to sustain life, maintaining low entropy levels. The brain, while maintaining lower entropy compared to an equilibrium state, shows variations in local entropy across different areas, reflecting functional differences [4, 5].
In this work our basis is the categorization reported in the Diagnostic and Statistical Manual of Mental Disorders, which defines six domains of cognition (Table 1), which are complex attention, executive function, learning and memory, language, perceptual-motor control and social cognition [6,7].
Through this systematic literature review, we explored several questions: which cognitive states have been studied using entropy analysis, the types of entropy used, and the consistency of findings across studies. We analyzed methodologies employed in both resting state and cognitive task research to assess the current state of the field and identify gaps in these approaches. Our aim is to understand future trends and potential developments in this area of research.

Materials and

Methods:  We used the Web of Science database as our main source of scientific articles. We searched for articles with "fMRI" and "entropy" in their topics within the 2000 - 2023 timeframe, limiting our search to journal articles, proceeding papers, review articles, meeting abstracts, and book chapters. This search resulted in 538 manuscripts.
We downloaded the manuscript information as an Excel file and applied four filters to create four tags (Figure 1):
1. Papers with "fMRI" and "entropy" in their titles
2. Papers with "fMRI" and "entropy" in the authors’ keywords
3. Papers with "fMRI" and "entropy" in the "Keywords Plus"
4. Papers with "fMRI" and "entropy" in their abstracts
We eliminated the papers that do not have any of these tags and created a list of articles.
Our study included only human subjects, excluding any animal studies and data simulation research. We thoroughly reviewed the intervention designs within the papers to determine the cognitive states of the subjects. We categorized various types of entropies into six classifications (Table 2, [8-13]): Approximate Entropy, Sample Entropy, Multiscale Entropy, Transfer Entropy, Shannon Entropy, and Fuzzy Entropy. Papers that employed entropy in their analysis without specifying the type were grouped under "Unspecified."

Results, Conclusions, and Discussions: We have identified 538 papers, 443 met the filtering criteria. We utilized these papers to explore the application of entropy in fMRI data analysis by addressing three key questions.
First, we examined whether entropy was used in studies involving healthy or sick subjects. Among the 443 papers, 60.9% examined subjects without psychological disorders. In contrast, 32.5% focused on subjects with psychological disorders, including schizophrenia, autism, Alzheimer's or mild cognitive impairment, attention deficit hyperactivity disorder, and multiple cognitive disorders. The remaining examined non-human subjects or signal simulations.
Second, we investigated which cognitive domains were studied using entropy. Among the papers researching human subjects, 55.31% focused on the resting state, 16.91% on perceptual-motor control, 7.25% on multiple cognitive tasks, 6.28% on learning and memory, 3.86% on social cognition, 3.14% on language, 3.14% on executive function, 2.17% on attention, and the rest were combination of variability of tasks.
Finally, we explored the types of entropy used in these studies. Among the papers 65.95% used Shannon Entropy (SE), 11.35% used Sample Entropy (SampEn), 4.32% used Transfer Entropy (TE), 2.70% used Multiscale Entropy (MSE), and 1.62% used Approximate Entropy (ApEn). Additionally, 0.54% used both Approximate Entropy and Sample Entropy, 0.54% used both Multiscale Entropy and Fuzzy Entropy, 0.54% used both Multiscale Entropy and Shannon Entropy, and the rest used complexity measures inspired by one or multiple entropy methods.
The results indicate a reduction in overall brain complexity due to aging and cognitive decline [14]. However, in specific cases like Alzheimer's disease, evidence shows an increase in overall brain entropy [14]. The primary gaps in the findings were in studies focused on unhealthy subjects and cognitive tasks. These gaps are due to variations in the health status of subjects in studies of unhealthy patients and the variability in intervention designs in studies of cognitive tasks. Additionally, most research has focused on overall brain entropy, with limited attention to the entropy of specific brain regions or networks.

Acknowledgements (Optional):  Reference:
[1] Biswal, B. B., Mennes, M., Zuo, X. N., Gohel, S., Kelly, C., Smith, S. M., et al. (2010). Toward discovery science of human brain function. Proc. Natl. Acad. Sci. U.S.A. 107, 4734–4739. doi: 10.1073/pnas.0911855107
[2] Nezafati, M., Temmar, H., & Keilholz, S. D. (2020). Functional MRI signal complexity analysis using sample entropy. Frontiers in neuroscience, 14, 520547.
[3] Schrödinger, E. (1945). What is Life? The Physical Aspect of the Living Cell. Cambridge: Cambridge University Press.
[4] Bergström, R. M. (1969). An entropy model of the developing brain. Dev. Psychobiol. 2, 139–152. doi: 10.1002/dev.420020304.
[5] Singer, W. (2009). The brain, a complex self-organizing system. Eur. Rev. 17, 321–329. doi: 10.1017/S10627987090007510.
[6] American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). https://doi.org/10.1176/appi.books.9780890425596.
[7] Sachdev, Perminder & Blacker, Deborah & Blazer, Dan & Ganguli, Mary & Jeste, Dilip & Paulsen, Jane & Petersen, Ronald. (2014). Classifying neurocognitive disorders: The DSM-5 approach. Nature Reviews Neurology. 10. 634-642. 10.1038/nrneurol.2014.181.
[8] Karaca, Y., & Moonis, M. (2022). Shannon entropy-based complexity quantification of nonlinear stochastic process. Multi-Chaos, Fractal and Multi-Fractional Artificial Intelligence of Different Complex Systems, 231–245. https://doi.org/10.1016/b978-0-323-90032-4.00018-3
[9] Richman, J. S., Lake, D. E., & Moorman, J. R. (2004). Sample entropy. Methods in Enzymology, 172–184. https://doi.org/10.1016/s0076-6879(04)84011-4
[10] Ursino, M., Ricci, G., & Magosso, E. (2020). Transfer entropy as a measure of brain connectivity: A critical analysis with the help of neural mass models. Frontiers in Computational Neuroscience, 14. https://doi.org/10.3389/fncom.2020.00045
[11] Costa, M., Goldberger, A. L., & Peng, C.-K. (2005). Multiscale entropy analysis of Biological Signals. Physical Review E, 71(2). https://doi.org/10.1103/physreve.71.021906
[12] Delgado-Bonal, A., & Marshak, A. (2019). Approximate Entropy and Sample Entropy: A Comprehensive Tutorial. Entropy (Basel, Switzerland), 21(6), 541. https://doi.org/10.3390/e21060541
[13] Al-Sharhan, Salah & Karray, Fakhri & Gueaieb, W. & Basir, Otman. (2001). Fuzzy entropy: A brief survey. IEEE International Conference on Fuzzy Systems. 3. 1135 - 1139. 10.1109/FUZZ.2001.1008855.
[14] Wang, Z., & Alzheimer’s Disease Neuroimaging Initiative. (2020). Brain entropy mapping in healthy aging and Alzheimer’s disease. Frontiers in Aging Neuroscience, 12, 596122.