Whenmakes you unique: temporality of the human brain fingerprint

Abstract

The extraction of fingerprints from human brain connectivity data has become a new frontier in neuroscience. However, the time scales of human brain identifiability have not been addressed yet. In other words, what temporal features make our brains more identifiable? We here explore the dynamics of brain fingerprints (or brainprints) along two complementary axes: 1) what is the optimal time scale at which brainprints integrate sufficient information, 2) when best identification happens. Using dynamic identifiability, we show that the best identification emerges at longer time scales (~300s); however, short transient bursts of identifiability persist even when looking at shorter functional interactions. We find that these bursts of identifiability might be strongly associated with neuronal activity. Furthermore, we report evidence that different parts of connectome fingerprints relate to different time scales: i.e., more visual-somatomotor at short temporal windows, more frontoparietal-DMN driven by increasing temporal windows. Finally, using a meta-analytic approach, we show that there is a broad spectrum of associations between brainprints and behavior. At faster time scales, human brain fingerprints are linked to multisensory stimulation, eye movements, affective processing, visuospatial attention. At slower time scales instead, we find higher-cognitive functions, such as language and verbal semantics, awareness, declarative and working memory, social cognition. We hope that this first investigation of the temporality of the human brain fingerprint will pave the way towards a better understanding of what and when makes our brains unique.

Brain fingerprints along the language hierarchy

“…The increased individual identifiability may be related to the fact that comprehending stories requires the integration of information over a longer time scale than did the two lower-level tasks. This is consistent with previous findings that the best identification emerges at longer time scales (Van De Ville et al, 2021).…”

“…The data were shifted back in time by 4 s to account for the hemodynamic lag of blood-oxygen-leveldependent (BOLD) signals. As previous studies have shown that data length can influence the accuracy of identification (Van De Ville et al, 2021), we extracted an equal number of time points (N = 32) from each session. This number was determined by the maximum data length of the backward and the scrambled conditions.…”

“…Currently, only a few studies have explored factors that potentially influence individual identifiability, including: (i) the temporal window used to compute FC profiles. It has been reported that the greater identifiability occurred at longer time scales (Van De Ville et al, 2021); (ii) the anatomical loci. Across the whole brain, the connectivity profiles of the frontoparietal network and medial frontal networks were most distinctive for individuals (Amico and Goni, 2018); and (iii) factors affecting fMRI data which might be unique to individuals and stable enough across time, including global signals (Chen and Hu, 2018), head motion and brain anatomy (Finn et al, 2015).…”

Fingerprints of brain disease: Connectome identifiability in cognitive decline and Alzheimer’s disease

“…2A]. Note that these ISelf values are equivalent to those found by existing studies in healthy subjects of the Human Connectome Project database 47 . Permutation testing showed that IDiff and Success-rate were different from null distributions at p <.001 in all groups (Fig.…”

“…Future studies should explore how fingerprints change across scanning sessions, or even longitudinally during cognitive decline (i.e., over days or months). It would be also interesting to further explore the temporality of brain fingerprints in neurodegeneration 47 .…”