Abstract
We applied detrended fluctuation analysis, power spectral density, and eigenanalysis of detrended cross-correlations to investigate fMRI data representing a diurnal variation of working memory in four visual tasks: two verbal and two nonverbal. We show that the degree of fractal scaling is regionally dependent on engagement in cognitive tasks. A particularly apparent difference was found between memorisation in verbal and nonverbal tasks. Furthermore, the detrended cross-correlations between brain areas were predominantly indicative of differences between resting state and other tasks, between memorisation and retrieval, and between verbal and nonverbal tasks. The fractal and spectral analyses presented in our study are consistent with previous research related to visuospatial and verbal information processing, working memory (encoding and retrieval), and executive functions, but they were found to be more sensitive than Pearson correlations and showed the potential to obtain other subtler results. We conclude that regionally dependent cognitive task engagement can be distinguished based on the fractal characteristics of BOLD signals and their detrended cross-correlation structure.
Introduction
Neuroscientists typically study working memory by mapping brain activation or calculating simple linear correlations between regions. However, Blood Oxygen Level Dependent (BOLD) signals from fMRI are complex and non-stationary, exhibiting long-range temporal correlations. This means the brain's activity follows a mathematically self-similar, or "fractal," pattern where past states influence future ones.
Standard tools like Pearson correlations often miss the subtle "texture" of these signals. This study investigates whether the fractal characteristics of brain activity change depending on the type of information being processed (verbal vs. non-verbal) and the specific stage of the memory process (encoding vs. retrieval).
Methods
The researchers analyzed fMRI data from 54 healthy participants performing four visual tasks: two verbal (semantic words and phonological pseudowords) and two non-verbal (global and local processing of abstract figures). The analysis pipeline focused on both temporal and cross-correlations to capture the hierarchical structure of brain activity.
As shown in Figure 1, the team employed several sophisticated techniques:
- Detrended Fluctuation Analysis (DFA): To calculate the Hurst exponent ($H$), measuring signal persistence.
- Power Spectral Density (PSD): To find the spectral exponent ($\beta$), confirming fractal scaling.
- Eigenanalysis of Detrended Cross-Correlations (DCC): To identify how networks synchronize while accounting for signal non-stationarity.
Key Results and Insights
The study revealed that the brain's "fractalness" is highly dynamic. During the resting state, the brain is most persistent ($H \approx 1.1$). However, when engaged in a task, the Hurst exponent drops significantly ($H \approx 0.7$), suggesting the signal becomes less predictable and more "efficient" for processing new information.
Encoding vs. Retrieval
A major finding was that the drop in $H$ was significantly more pronounced during the encoding phase (memorizing) than the retrieval phase (recalling), particularly in the visual cortex (see Figure 3). This suggests that "writing" information into working memory requires a more complex, less persistent neural state than "reading" it back out.
Task-Specific Signatures
The analysis also distinguished between verbal and non-verbal processing. Verbal tasks showed lower $H$ values in sensory-motor and visual regions compared to non-verbal tasks. These fractal metrics proved more sensitive than traditional Pearson correlations, which often failed to distinguish between the subtle differences of encoding and retrieval phases.
Conclusion
Regionally dependent cognitive engagement can be accurately distinguished based on the fractal characteristics of BOLD signals. While fMRI has limitations in temporal resolution, these fractal and spectral analyses offer a more nuanced view of how the brain reorganizes its temporal processing patterns to meet specific cognitive demands. Future work may apply these methods to higher-frequency data like EEG to further explore these dynamic neural "fingerprints."
Reference: Ochab, J. K., et al. (2022). Task-dependent fractal patterns of information processing in working memory. arXiv:2209.13297v1 [q-bio.NC].
Summary updated