Meso-scale network analysis of resting state-fMRI brain network connectivity performs poorly as a prognostic tool in critically ill traumatic brain injury patients
Prognostication is important in the management of critically ill traumatic brain injury (TBI) patients, as it may guide decisions on treatment intensity. Changes in functional connectivity that arise due to TBI have been observed in several brain networks using resting-state functional magnetic reso...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2022-03-01
|
| Series: | Neuroimage: Reports |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2666956022000034 |
| _version_ | 1828864230350127104 |
|---|---|
| author | Jonathan Tjerkaski William H. Thompson Bo-Michael Bellander Eric P. Thelin Peter Fransson |
| author_facet | Jonathan Tjerkaski William H. Thompson Bo-Michael Bellander Eric P. Thelin Peter Fransson |
| author_sort | Jonathan Tjerkaski |
| collection | DOAJ |
| description | Prognostication is important in the management of critically ill traumatic brain injury (TBI) patients, as it may guide decisions on treatment intensity. Changes in functional connectivity that arise due to TBI have been observed in several brain networks using resting-state functional magnetic resonance imaging (rs-fMRI). Studies suggest that cognitive dysfunction following TBI can, in part, be explained by changes in functional connectivity. The overall aim of this study was to investigate whether easy-to-calculate graph theory-based summaries of brain network functional connectivity can discriminate between patients with favorable and unfavorable outcomes. To further substantiate these results, we used the Louvain algorithm to detect the community structure of each TBI patient, to assess whether it differs from that of healthy individuals.We performed fMRI on 44 TBI patients who were admitted to the neurocritical care unit. Long-term functional outcome was determined at 6–12 months using the Glasgow outcome scale (GOS), with GOS 1–3 being defined as an unfavorable outcome. Functional connectivity was measured using Pearson’s correlation coefficient. We derived topographical summaries of brain network connectivity using graph theory. We computed the participation coefficient (PC) and the module degree z-score (MDZ) for each parcel in the cortical template of Thomas Yeo et al. (2011). As topographical summaries of the network parameters we used the median, maximum and standard deviation (SD) of PC and MDZ in each of the seven functional networks that are included in the cortical parcellation of Yeo et al. Age and sedative medications were used as nuisance regressors.Using multivariate logistic regression, we observed no statistically significant association between the topographical summaries of brain connectivity and the dichotomized GOS. Thus, although graph theoretical measures applied to brain network connectivity patterns have previously been reported to be associated with cognitive dysfunction, they appear to be unable to predict long-term functional outcomes in critically ill TBI patients. Similarly, the extent to which TBI patient’s community structures differed from that of healthy individuals, measured using adjusted mutual information, was also not associated with the dichotomized GOS.In summary, our results suggest that Meso-scale graph theoretical analyses of rs-fMRI brain network connectivity patterns can be expected to perform poorly as a prognostic tool in critically ill TBI patients. |
| first_indexed | 2024-12-13T04:02:08Z |
| format | Article |
| id | doaj.art-0a4985da45f04dd5a03da2ebf679db7b |
| institution | Directory Open Access Journal |
| issn | 2666-9560 |
| language | English |
| last_indexed | 2024-12-13T04:02:08Z |
| publishDate | 2022-03-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Neuroimage: Reports |
| spelling | doaj.art-0a4985da45f04dd5a03da2ebf679db7b2022-12-22T00:00:25ZengElsevierNeuroimage: Reports2666-95602022-03-0121100079Meso-scale network analysis of resting state-fMRI brain network connectivity performs poorly as a prognostic tool in critically ill traumatic brain injury patientsJonathan Tjerkaski0William H. Thompson1Bo-Michael Bellander2Eric P. Thelin3Peter Fransson4Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden; Corresponding author.Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, SwedenDepartment of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden; Department of Neurosurgery, Karolinska Institutet, Stockholm, SwedenDepartment of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden; Department of Neurology, Karolinska University Hospital, Stockholm, SwedenDepartment of Clinical Neuroscience, Karolinska Institutet, Stockholm, SwedenPrognostication is important in the management of critically ill traumatic brain injury (TBI) patients, as it may guide decisions on treatment intensity. Changes in functional connectivity that arise due to TBI have been observed in several brain networks using resting-state functional magnetic resonance imaging (rs-fMRI). Studies suggest that cognitive dysfunction following TBI can, in part, be explained by changes in functional connectivity. The overall aim of this study was to investigate whether easy-to-calculate graph theory-based summaries of brain network functional connectivity can discriminate between patients with favorable and unfavorable outcomes. To further substantiate these results, we used the Louvain algorithm to detect the community structure of each TBI patient, to assess whether it differs from that of healthy individuals.We performed fMRI on 44 TBI patients who were admitted to the neurocritical care unit. Long-term functional outcome was determined at 6–12 months using the Glasgow outcome scale (GOS), with GOS 1–3 being defined as an unfavorable outcome. Functional connectivity was measured using Pearson’s correlation coefficient. We derived topographical summaries of brain network connectivity using graph theory. We computed the participation coefficient (PC) and the module degree z-score (MDZ) for each parcel in the cortical template of Thomas Yeo et al. (2011). As topographical summaries of the network parameters we used the median, maximum and standard deviation (SD) of PC and MDZ in each of the seven functional networks that are included in the cortical parcellation of Yeo et al. Age and sedative medications were used as nuisance regressors.Using multivariate logistic regression, we observed no statistically significant association between the topographical summaries of brain connectivity and the dichotomized GOS. Thus, although graph theoretical measures applied to brain network connectivity patterns have previously been reported to be associated with cognitive dysfunction, they appear to be unable to predict long-term functional outcomes in critically ill TBI patients. Similarly, the extent to which TBI patient’s community structures differed from that of healthy individuals, measured using adjusted mutual information, was also not associated with the dichotomized GOS.In summary, our results suggest that Meso-scale graph theoretical analyses of rs-fMRI brain network connectivity patterns can be expected to perform poorly as a prognostic tool in critically ill TBI patients.http://www.sciencedirect.com/science/article/pii/S2666956022000034Functional magnetic resonance imagingTraumatic brain injuryGraph theoryPrognostication |
| spellingShingle | Jonathan Tjerkaski William H. Thompson Bo-Michael Bellander Eric P. Thelin Peter Fransson Meso-scale network analysis of resting state-fMRI brain network connectivity performs poorly as a prognostic tool in critically ill traumatic brain injury patients Neuroimage: Reports Functional magnetic resonance imaging Traumatic brain injury Graph theory Prognostication |
| title | Meso-scale network analysis of resting state-fMRI brain network connectivity performs poorly as a prognostic tool in critically ill traumatic brain injury patients |
| title_full | Meso-scale network analysis of resting state-fMRI brain network connectivity performs poorly as a prognostic tool in critically ill traumatic brain injury patients |
| title_fullStr | Meso-scale network analysis of resting state-fMRI brain network connectivity performs poorly as a prognostic tool in critically ill traumatic brain injury patients |
| title_full_unstemmed | Meso-scale network analysis of resting state-fMRI brain network connectivity performs poorly as a prognostic tool in critically ill traumatic brain injury patients |
| title_short | Meso-scale network analysis of resting state-fMRI brain network connectivity performs poorly as a prognostic tool in critically ill traumatic brain injury patients |
| title_sort | meso scale network analysis of resting state fmri brain network connectivity performs poorly as a prognostic tool in critically ill traumatic brain injury patients |
| topic | Functional magnetic resonance imaging Traumatic brain injury Graph theory Prognostication |
| url | http://www.sciencedirect.com/science/article/pii/S2666956022000034 |
| work_keys_str_mv | AT jonathantjerkaski mesoscalenetworkanalysisofrestingstatefmribrainnetworkconnectivityperformspoorlyasaprognostictoolincriticallyilltraumaticbraininjurypatients AT williamhthompson mesoscalenetworkanalysisofrestingstatefmribrainnetworkconnectivityperformspoorlyasaprognostictoolincriticallyilltraumaticbraininjurypatients AT bomichaelbellander mesoscalenetworkanalysisofrestingstatefmribrainnetworkconnectivityperformspoorlyasaprognostictoolincriticallyilltraumaticbraininjurypatients AT ericpthelin mesoscalenetworkanalysisofrestingstatefmribrainnetworkconnectivityperformspoorlyasaprognostictoolincriticallyilltraumaticbraininjurypatients AT peterfransson mesoscalenetworkanalysisofrestingstatefmribrainnetworkconnectivityperformspoorlyasaprognostictoolincriticallyilltraumaticbraininjurypatients |