Disrupted Information Flow in Resting-State in Adolescents With Sports Related Concussion

Abstract

Children and youths are at a greater risk of concussions than adults, and once injured, take longer to recover. A key feature of concussion is an increase in functional connectivity, yet it remains unclear how changes in functional connectivity relate to the patterns of information flow within resting state networks following concussion and how these relate to brain function. We applied a data-driven measure of directed effective brain connectivity to compare the patterns of information flow in healthy adolescents and adolescents with subacute concussion during the resting state condition. Data from 32 healthy adolescents (mean age =16 years) and 21 concussed adolescents (mean age = 15 years) within 1 week of injury were included in the study. Five minutes of resting state data EEG were collected while participants sat quietly with their eyes closed. We applied the information flow rate to measure the transfer of information between the EEG time series of each individual at different source locations, and therefore between different brain regions. Based on the ensemble means of the magnitude of normalized information flow rate, our analysis shows that the dominant nexus of information flow in healthy adolescents is primarily left lateralized and anterior-centric, characterized by strong bidirectional information exchange between the frontal regions, and between the frontal and the central/temporal regions. In contrast, adolescents with concussion show distinct differences in information flow marked by a more left-right symmetrical, albeit still primarily anterior-centric, pattern of connections, diminished activity along the central-parietal midline axis, and the emergence of inter-hemispheric connections between the left and right frontal and the left and right temporal regions of the brain. We also find that the statistical distribution of the normalized information flow rates in each group (control and concussed) is significantly different. This paper is the first to describe the characteristics of the source space information flow and the effective connectivity patterns between brain regions in healthy adolescents in juxtaposition with the altered spatial pattern of information flow in adolescents with concussion, statistically Hristopulos et al. Disrupted Information Flow in Concussed Adolescentsquantifying the differences in the distribution of the information flow rate between the two populations. We hypothesize that the observed changes in information flow in the concussed group indicate functional reorganization of resting state networks in response to brain injury.

The Effects of Mild Traumatic Brain Injury on Cognitive-Motor Integration for Skilled Performance

“…We propose that widespread injury can reduce rule-based skilled performance through a failure to communicate between brain networks. Cognitive brain network alterations have been observed following concussion (13)(14)(15)(16)(17). To assess the communication between brain areas for movement control, our group developed a touch screen based functional assessment tool using a visuomotor task that incorporates two forms of non-standard mapping: explicit rule integration (strategic control), and implicit spatial vision-to-hand-motion realignment (sensorimotor recalibration).…”

Recurrent Neural Network-based Acute Concussion Classifier using Raw Resting State EEG Data

“…However, prior to extracting the features, we had filtered the data [82][83][84] to remove signal drift, an approach that is commonly adopted to mitigate weak non-stationary behaviour. This works well enough when one is interested in population-level differences between concussed and healthy groups [24,28,47,48]. However, the distribution functions of features for the healthy and concussed individuals overlap, which hampers attempts to use the features to assess individual subjects.…”

“…Symptoms, however, do not appear to be directly correlated to the patho-physiological mechanisms responsible for the cognitive, emotional and behavioral deficits [13,14] and several studies [14][15][16][17][18][19][20][21][22][23] have shown that structural and functional abnormalities in the brain persist even after symptom resolution, especially in pediatric patients. Manning et al [20] and Hristopulos et al [24] hypothesize that the increased vulnerability of children and youths to brain injury, as well as their longer recovery time, are consequences of the effects of concussion being overlaid on brains whose structural and functional organization are undergoing dynamic changes due to development. Moreover, recovering individuals are also at a higher risk of sustaining additional concussions, which can lead to neuro-psychiatric and neuro-degenerative disorders [25,26].…”

“…At the group-level, structural changes in the integrity of white matter in adolescents has been observed using diffusion tensor imaging (DTI) [20,[27][28][29][30][31][32]. Similarly, studies of youths during "resting state" reveal significant mTBI-induced alterations in the functional organization of the brain with respect to their healthy counterparts, with key features being (i) a shift in spectral profile toward higher frequencies in the frontal brain regions, [33] (ii) an increase in functional connectivity (hyperconnectivity) [17,34,35], and (iii) disrupted information flow patterns [24], again particularly in the frontal regions. All of these studies rely on summary measures to characterize functional changes in the brain.…”

“…We explain our decision to use 90-second long samples in §4. Our decision to select EEG as the modality of choice was informed by a number of considerations, including the fact that (i) it is an inexpensive, non-invasive, direct, fine time resolution probe of the neural activity in individuals; (ii) at the group-level, it can reliably and objectively detect functional changes in the brain due to mTBI [24,28,47,48]; and (iii) it does not require elaborate infrastructure, the introduction of contrast agents and radioisotopes, or exposure to radiation. Additionally, resting state data is easy to collect; the participant is neither required to perform a specific task nor subjected to a stimulus.…”