Skip to main content
NCBI home page
NeuroImage: Clinical logoLink to NeuroImage: Clinical
. 2023 Mar 1;37:103361. doi: 10.1016/j.nicl.2023.103361

Default mode network overshadow executive control network in coma emergence and awakening prediction of patients with sTBI

Rui-zhe Zheng a,b,1, Can-xin Xu c,d, Lai-yang Zhou e, Dong-fu Feng f,g,⁎
PMCID: PMC9995458  PMID: 36871404

Highlights

  • •

    It is unclear the importance of DMN and ECN in acute coma.

  • •

    It is unclear whether the DMN and ECN can be used to predict awakening.

  • •

    The DMN plays more important role than the ECN and the DMN-ECN interaction in the traumatic coma.

  • •

    The DMN plays severe role than the ECN and the DMN-ECN interaction in awakening prediction.

Keywords: Severe traumatic brain injury, Coma, Default mode network, Executive control network, Prediction

Abstract

Objective

We aimed to explore the pathogenesis of traumatic coma related to functional connectivity (FC) within the default mode network (DMN), within the executive control network (ECN) and between the DMN and ECN and to investigate its capacity for predicting awakening.

Methods

We carried out resting-state functional magnetic resonance imaging (fMRI) examinations on 28 traumatic coma patients and 28 age-matched healthy controls. DMN and ECN nodes were split into regions of interest (ROIs), and node-to-node FC analysis was conducted on individual participants. To identify coma pathogenesis, we compared the pairwise FC differences between coma patients and healthy controls. Meanwhile, we divided the traumatic coma patients into different subgroups based on their clinical outcome scores at 6 months postinjury. Considering the awakening prediction, we calculated the area under the curve (AUC) to evaluate the predictive ability of changed FC pairs.

Results

We found a massive pairwise FC alteration in the patients with traumatic coma compared to the healthy controls [45% (33/74) pairwise FC located in the DMN, 27% (20/74) pairwise FC located in the ECN, and 28% (21/74) pairwise FC located between the DMN and ECN]. Moreover, in the awake and coma groups, there were 67% (12/18) pairwise FC alterations located in the DMN and 33% (6/18) pairwise FC alterations located between the DMN and ECN. We also indicated that pairwise FC that showed a predictive value of 6-month awakening was mainly located in the DMN rather than in the ECN. Specifically, decreased FC between the right superior frontal gyrus and right parahippocampal gyrus (in the DMN) showed the highest predictive ability (AUC = 0.827).

Conclusion

In the acute phase of severe traumatic brain injury (sTBI), the DMN plays a more prominent role than the ECN and the DMN-ECN interaction in the emergence of traumatic coma and the prediction of 6-month awakening.

1. Introduction

(Bayne et al., 2016)Severe traumatic brain injury (sTBI) is a leading cause of long-term neurological deficits and the primary cause in patients with prolonged coma. Conventional wisdom for determining the outcomes of patients with sTBI highlights the important role of predictive models.(Bayne et al., 2016, Bianciardi et al., 2021) However, an accurate prognostic model to predict the awakening trajectory following traumatic coma is still lacking because the consciousness mechanism has not been fully elucidated.

Recent functional magnetic resonance imaging (fMRI) studies have revealed that the brain activity of intrinsic networks may be important for supporting human consciousness.(Bodien et al., 2017) Generally, the default mode network (DMN) is an “intrinsic” system that is dedicated to oriented conscious processes, while the executive control network (ECN) is an “extrinsic” system that is involved in cognitive perception processes.(Boveroux et al., 2010) The relationship between DMN connectivity and consciousness has been established for a broad range of physiological and pathological conditions.(Crone et al., 2011) Meanwhile, ECN connectivity has been found to be involved in the voluntary control of consciousness behavior and drug-induced consciousness alteration.(Davis et al., 2007, Edlow et al., 2017) Subsequent proposals indicated that both the DMN and ECN together constitute the frontoparietal network, with functional connectivity (FC) increasing as consciousness improves following brain injury.(Edlow et al., 2021) Given that, a good question is whether the DMN, the ECN or the DMN-ECN interaction is more important in coma pathogenesis.

Early identification of the degree of brain function deficiency may be useful not only in revealing the severity of consciousness impairment but also in predicting the potential of consciousness recovery.(Fransson, 2005, Fridman et al., 2014) However, the abnormal connectome of typical cortical networks (i.e., DMN and ECN) in patients with traumatic coma has not been well established to date. These gaps in knowledge have profound clinical implications for clinical decision making in patients with sTBI. For example, without knowing whether a patient can recover from coma, many hopeless families voluntarily withdraw life-sustaining treatment; these decisions account for up to 68% of deaths in patients with sTBI.(He et al., 2015, Izzy et al., 2013).

The purpose of the present study was to identify alterations in two large-scale functional networks (DMN and ECN) that are compatible with acute coma and associated with consciousness recovery in patients with sTBI. Specifically, in this work, we had two aims: 1) to establish which networks showed larger alterations in traumatic coma and 2) to investigate which network FCs could be better used for predicting coma recovery.

2. Methods

2.1. Participant selection

The present study included sTBI patients who were ≥ 18 years old who did not have basic diseases before brain injury. Importantly, the inclusion criteria were further restricted to closed craniocerebral injury with no surgical operation. We defined the criteria of traumatic coma in this study as patients who presented with an initial Glasgow Coma Scale (GCS) score ≤ 8 points, with no eye-opening or verbal response for at least 24 h on examinations unconfounded by sedation or paralysis13. The exclusion criteria included patients with bilateral dilated and fixed pupils, circulatory failure, pregnancy, comorbidities other neurological and/or psychiatric diseases, alcohol and drug abuse, and extracranial body regions with Abbreviated Injury Scale > 2. Baseline characteristics-matched healthy subjects were also recruited as the healthy controls.

The severity of the initial brain injury was assessed by the Glasgow Outcome Scale (GCS), and the outcome at 6 months postinjury was evaluated by the Glasgow Outcome Scale - Extended (GOSE). In this study, we used two indicators to reflect the outcome of TBI patients. First, we used the command-following behavior as the primary outcome (Outcome I) measure for recovery from coma because it is a commonly used measure of responsiveness in studies of patients with acute sTBI8,(Leblanc et al., 2018). Second, we used the GOSE score as the secondary outcome (Outcome II) measure, which can be dichotomized into poor (GOSE 1–4) and good (GOSE 5–8) outcomes15.

2.2. MRI data acquisition

The blood oxygen level-dependent (BOLD) fMRI data were acquired with an 8-channel standard quadrature head coil on a Discovery MR750 3 T MRI scanner (GE Medical Systems, Milwaukee, WI, USA) using the following parameters: percent phase field view = 100, flip angle = 90°, TE = 30 ms, TR = 2000 ms, FOV = 300 × 300 mm; matrix = 64 × 64, thickness = 3.5 mm, spacing between slices = 4.2 mm, and slices = 33, 8 min. Meanwhile, whole-brain high-spatial resolution 3D T1-weighted images were obtained for coregistration and anatomical location purposes using the following pulse sequence: TE = 3.18 ms, TR = 8.16 ms, FOV = 256 × 256 mm, flip angle = 12°, matrix = 256 × 256, slice thickness = 1 mm, interslice gap = 0 mm, and 168 slices.

For patients with acute traumatic coma, the MRI scan was performed as soon (approximately 2 weeks after admission) as the patient was clinically stable for transport to the scanner, as determined and accompanied by full-time neurosurgeons with specific training in critical care. All patients who received drug sedation (e.g., benzodiazepine or opioid) underwent a washout period for at least 24 h before the MRI scan. Meanwhile, the MRI data for healthy controls were acquired with the same scanning procedures.

2.3. Data preprocessing

All fMRI data preprocessing was performed using MATLAB 2016b. The DPABI toolbox (https://rfmri.org/DPABI) was used for the procedure, which included remove first (10 time series), slice timing, segment, head-motion correction, and spatial normalization. Significantly, the patients with sTBI were inevitably restless in the acute phase, and participants with more than a 2.0 mm maximum translation and 2 degrees of rotation (x, y, and z directions) were excluded. Then, aFriston 24 and Voxel-specific 12 model was applied to regress out head motion effects. After that, linear regression was used to remove spurious covariates along with temporal derivatives (such as white matter and cerebrospinal fluid signals). Then, the low-frequency drift, physiologically high-frequency (respiratory and cardiac noise), and time-series linear detrending were reduced through bandpass filtration (0.01–0.1 Hz) and linearly detrended. In addition, the fMRI images were normalized to the Montreal Neurological Institute (MNI) template and spatially resampled at a resolution of 3 × 3 × 3 mm3 space. Finally, the data were smoothed using a Gaussian kernel of 4 × 4 × 4 mm3 full width at half-maximum (FWHM).

2.4. Network regions of interest construction

The regions of interest (ROIs) reference canonical resting state network templates, which were provided by Stanford’s Functional Imaging in Neuropsychiatric Disorders (FIND) Laboratory.(Mason et al., 2007) The network ROI-nodes were constructed with reference to the independent components identified and labeled by their corresponding networks. As shown in Fig. 1, a total of 31 ROI nodes that belong to the DMN (ROIs = 19) and the ECN (ROIs = 12) were retained for subsequent ROIwise analysis (for detailed MNI space information, see Supplementary Table S1).

Fig. 1.

Fig. 1

Open in a new tab

Schematic visualization of regions of interest in default mode network and executive control network [A. dorsal default mode network; B. ventral default mode network; C. left executive control network; D. right executive control network (Brain 0: medial frontal gyrus, Brain 1: left angular gyrus, Brain 2: right superior frontal gyrus, Brain 3: posterior cingulate, Brain 4: median cingulate and paracingulate gyri, Brain 5: right supramarginal gyrus, Brain 6: thalamus, Brain 7: left parahippocampagyruss, Brain 8: right parahippocampagyruss, Brain 9: left posterior cingulate, Brain 10: left superior frontal gyrus, Brain 11: left fusiform gyrus, Brain 12: left middle occipital gyrus, Brain 13: right posterior cingulate, Brain 14: superior parietal lobule, Brain 15: right superior frontal gyrus, Brain 16: right fusiform gyrus, Brain 17: right middle occipital gyrus, Brain 18: right cerebellar tonsil, Brain 19: left dorsolateral prefrontal cortex; Brain 20: left middle frontal gyrus, Brain 21: left inferior parietal lobule, Brain 22: middle temporal gyrus, Brain 23: cerebellum posterior lobe, Brain 24: left thalamus, Brain 25: right dorsolateral prefrontal cortex, Brain 26: right middle frontal gyrus, Brain 27: right inferior parietal lobule, Brain 28: right superior frontal gyrus, medial, Brain 29: left cerebellum posterior lobe, Brain 30: right thalamus)].

2.5. fMRI data analysis

The DPABI toolbox (https://rfmri.org/DPABI) was applied for the ROI-wise analysis. The time series was extracted from each ROI listed above (Fig. 1 and Supplement Table S1). Then, the FC matrix (31 × 31) was constructed by Pearson correlations with time series, and the results of correlation coefficients were Fisher transformed to Z values to increase normality.

The two-sample one-tailed t test was used to assess the group differences (coma patients vs. healthy controls; wake group vs. coma group) for the interregional connections. Multiple linear regression was used to adjust the effects of age and gender. We utilized a network-based statistic (NBS) approach in the GRETNA toolbox (https://www.nitrc.org/projects/gretna/) to identify the specific ROI-node pairs in which FC was altered.(Qin and Northoff, 2011) We calculated a corrected p value using the null distribution of the maximal connected component size, which was derived using a nonparametric permutation approach (5000 permutations). The statistical significance was set as p < 0.05.

2.6. Statistical analysis

Normally distributed data (Kolmogorov-Smirnov test) are represented by the mean ± standard deviation. Two-tailed two-sample t tests (measurement data) and chi-square tests (enumeration data) were applied to the transverse analysis and longitudinal analysis (SPSS 26.0; SPSS, Inc., Chicago, IL, United States). To test the clinical utility of fMRI parameters, that is, whether FC alterations in the DMN or ECN and between the DMN and ECN can predict the outcome or awakening of sTBI patients at 6 months. We constructed receiver operating characteristic (ROC) curves to test their predictive value. The area under the ROC curve (AUC) was also calculated to evaluate the discriminative ability of these patterns. An AUC of 0.5 implied no discriminative ability, whereas an AUC of 1.0 implied perfect discrimination. The significance level was set at p < 0.05.

3. Results

3.1. Demographics and characteristics of participants

After obtaining permission from the institutional ethics committee and written informed consent from surrogate decision-makers, we eventually included a total of 28 patients with traumatic coma and 28 baseline feature-matched healthy controls in this study (Fig. 2). The demographic and clinical characteristics of the participants and their outcome information are summarized in Table 1.

Fig. 2.

Fig. 2

Open in a new tab

Flowchart of the study. (n, number of participants; GCS, Glasgow Coma Scale; fMRI, functional magnetic resonance imaging).

Table 1.

Baseline characteristics of the participants.

Terms
Subjects 		Age
(Mean ± SD) 		Gender
male (%) 		Right-handed
N (%) 		aGCS
(Mean ± SD) 		GOS-E
(Mean ± SD)
All
Healthy controls 55.21 ± 18.09 18 (64.3) 28 (100.0) – –
Coma patients 58.00 ± 15.39 19 (67.9) 28 (100.0) 5.00 ± 2.45 3.00 ± 1.93
Outcome I
Good outcome group 64.38 ± 14.91 5 (62.5) 8 (100.0) 6.13 ± 2.23 5.50 ± 1.20
Poor outcome group 54.95 ± 15.09 14 (70.0) 20 (100.0) 4.55 ± 2.43 2.00 ± 1.03
Outcome II
Wake group 52.23 ± 17.22 7 (87.5) 13 (100.0) 5.85 ± 2.38 4.69 ± 1.49
Coma group 59.38 ± 15.98 8 (61.5) 8 (100.0) 2.88 ± 1.36 2.00 ± 0.00
Statistical results
P value .54a;.15b;.36c .78a;.70b;.20c – .13b;.002c <.001b, c
Open in a new tab

SD: Standard error; aGCS: admission Glasgow coma score; GOS-E: Glasgow outcome score expansion (a. healthy controls vs. coma patients; b. good outcome group vs. poor outcome group; c. wake group vs. coma group).

3.2. Pairwise FC alterations in the DMN and ECN and between the DMN and ECN

Compared with the healthy controls, there were significant differences in 74 pairwise FC values in sTBI patients who presented with acute coma (p < 0.05, NBS corrected), among which 45% of pairwise nodes (33/74) were located in the DMN, 27% of pairwise nodes (20/74) were located in the ECN, and 28% of pairwise nodes (21/74) were located between the DMN and ECN. More specifically, 96% (71/74) of pairwise nodes (edges) in and across these two networks showed decreased FC, while only 4% (3/74) showed increased FC in patients with traumatic coma. In terms of the nodes with decreased pairwise FC, most of them were in the DMN rather than the ECN as well as the DMN-ECN interaction (Fig. 3). These findings may demonstrate that the DMN plays a more important role than the ECN and the DMN-ECN interaction in the maintenance of consciousness.

Fig. 3.

Fig. 3

Open in a new tab

Differences of the paired functional connectivity within and across default mode network and executive control network between coma patients and health controls [A. T statistic matrix; B. P value matrix; C. brain regions with increased functional connectivity; D. regions of interest with decreased functional connectivity (Brain 0: medial frontal gyrus, Brain 1: left angular gyrus, Brain 2: right superior frontal gyrus, Brain 3: posterior cingulate, Brain 4: median cingulate and paracingulate gyri, Brain 5: right supramarginal gyrus, Brain 6: thalamus, Brain 7: left parahippocampagyruss, Brain 8: right parahippocampagyruss, Brain 9: left posterior cingulate, Brain 10: left superior frontal gyrus, Brain 11: left fusiform gyrus, Brain 12: left middle occipital gyrus, Brain 13: right posterior cingulate, Brain 14: superior parietal lobule, Brain 15: right superior frontal gyrus, Brain 16: right fusiform gyrus, Brain 17: right middle occipital gyrus, Brain 18: right cerebellar tonsil, Brain 19: left dorsolateral prefrontal cortex; Brain 20: left middle frontal gyrus, Brain 21: left inferior parietal lobule, Brain 22: middle temporal gyrus, Brain 23: cerebellum posterior lobe, Brain 24: left thalamus, Brain 25: right dorsolateral prefrontal cortex, Brain 26: right middle frontal gyrus, Brain 27: right inferior parietal lobule, Brain 28: right superior frontal gyrus, medial, Brain 29: left cerebellum posterior lobe, Brain 30: right thalamus)].

3.3. Differences in pairwise FC in the coma and awake

We further investigated whether the pairwise FC alterations in the DMN, the ECN, and the DMN-ECN interaction were compatible with neurological functional and coma recovery. First, we split all 28 patients into good (n = 8) and poor outcome groups (n = 20) according to their GOS-E scores. We found that there was no difference in pairwise FC between the good and poor outcome groups (Fig. 4). Seven patients were excluded from the study cohort due to death during follow-up. Then we split the remaining 21 patients were divided into coma (n = 8) and awake (n = 13) groups. Significantly, there were 67% (12/18) pairwise FC differences in the DMN and 33% (6/18) pairwise FC differences in the DMN-ECN interaction in the coma and awake groups. However, we did not find a statistically significant difference in pairwise FC in the ECN between groups (Fig. 5).

Fig. 4.

Fig. 4

Open in a new tab

Between-group differences in paired functional connectivity within and across networks in the good and poor outcome groups [A. T statistic matrix; B. P value matrix (Brain 0: medial frontal gyrus, Brain 1: left angular gyrus, Brain 2: right superior frontal gyrus, Brain 3: posterior cingulate, Brain 4: median cingulate and paracingulate gyri, Brain 5: right supramarginal gyrus, Brain 6: thalamus, Brain 7: left parahippocampus, Brain 8: right parahippocampus, Brain 9: left posterior cingulate, Brain 10: left superior frontal gyrus, Brain 11: left fusiform gyrus, Brain 12: left middle occipital gyrus, Brain 13: right posterior cingulate, Brain 14: superior parietal lobule, Brain 15: right superior frontal gyrus, Brain 16: right fusiform gyrus, Brain 17: right middle occipital gyrus, Brain 18: right cerebellar tonsil, Brain 19: left dorsolateral prefrontal cortex; Brain 20: left middle frontal gyrus, Brain 21: left inferior parietal lobule, Brain 22: middle temporal gyrus, Brain 23: cerebellum posterior lobe, Brain 24: left thalamus, Brain 25: right dorsolateral prefrontal cortex, Brain 26: right middle frontal gyrus, Brain 27: right inferior parietal lobule, Brain 28: right superior frontal gyrus, medial, Brain 29: left cerebellum posterior lobe, Brain 30: right thalamus)].

Fig. 5.

Fig. 5

Open in a new tab

Paired functional connectivity differences within and across networks between coma and wake groups [A. T statistic matrix; B. regions of interest with increased functional connectivity; C. regions of interest with decreased functional connectivity; D. P value matrix; E. quantitative comparison of altered functional connectivity (Brain 0: medial frontal gyrus, Brain 1: left angular gyrus, Brain 2: right superior frontal gyrus, Brain 3: posterior cingulate, Brain 4: median cingulate and paracingulate gyri, Brain 5: right supramarginal gyrus, Brain 6: thalamus, Brain 7: left parahippocampagyruss, Brain 8: right parahippocampagyruss, Brain 9: left posterior cingulate, Brain 10: left superior frontal gyrus, Brain 11: left fusiform gyrus, Brain 12: left middle occipital gyrus, Brain 13: right posterior cingulate, Brain 14: superior parietal lobule, Brain 15: right superior frontal gyrus, Brain 16: right fusiform gyrus, Brain 17: right middle occipital gyrus, Brain 18: right cerebellar tonsil, Brain 19: left dorsolateral prefrontal cortex; Brain 20: left middle frontal gyrus, Brain 21: left inferior parietal lobule, Brain 22: middle temporal gyrus, Brain 23: cerebellum posterior lobe, Brain 24: left thalamus, Brain 25: right dorsolateral prefrontal cortex, Brain 26: right middle frontal gyrus, Brain 27: right inferior parietal lobule, Brain 28: right superior frontal gyrus, medial, Brain 29: left cerebellum posterior lobe, Brain 30: right thalamus; * or#represents P < 0.05)].

3.4. Early detection of abnormal FC could predict awakening

The differences in FC were clarified between the coma and awake groups. Then, we tested these roles in predicting awakening. Our findings suggested that altered pairwise FC could be used to predict patients’ awakening for 6 months. Of note, the altered FC pairs were mainly located in the DMN and in the DMN-ECN rather than the ECN. Meanwhile, our results revealed that not only increased FC but also decreased FC yielded roughly equal predictive ability for predicting coma recovery. Specifically, decreased FC between the right superior frontal gyrus and right parahippocampal gyrus located within the DMN showed the highest predictive ability (Fig. 6.). In short, these findings revealed that the DMN and the DMN-ECN interaction play more important roles than the ECN in the prediction of 6-month awakening (for detailed AUC information, see Supplementary Table S2).

Fig. 6.

Fig. 6

Open in a new tab

Discriminative capacity of functional connectivity strength (Z value) for coma recovery (Brain 8 to 4: functional connectivity between right parahippocampagyruss and median cingulate paracingulate gyri; Brain 11 to 8: functional connectivity between left fusiform gyrus and right parahippocampagyruss; Brain 12 to 1: functional connectivity between left middle occipital gyrus and left angular gyrus; Brain 12 to 5: functional connectivity between left middle occipital gyrus and right supramarginal gyrus; Brain 15 to 6: functional connectivity between right superior frontal gyrus and thalamus; Brain 15 to 8: functional connectivity between right superior frontal gyrus and right parahippocampagyruss; Brain 16 to 8: functional connectivity between right fusiform gyrus and right fusiform gyrus; Brain 16 to 15: functional connectivity between right fusiform gyrus and right superior frontal gyrus; Brain 18 to 7: functional connectivity between right cerebellar tonsil and left parahippocampagyruss; Brain 18 to 8: functional connectivity between right cerebellar tonsil and right parahippocampagyruss; Brain 18 to 11: functional connectivity between right cerebellar tonsil and left fusiform gyrus; Brain 18 to 16: functional connectivity between right cerebellar tonsil and right fusiform gyrus; Brain 19 to 16: functional connectivity between left dorsolateral prefrontal cortex and right fusiform gyrus; Brain 24 to 7: functional connectivity between left thalamus and left parahippocampagyruss; Brain 24 to 8: functional connectivity between left thalamus and right parahippocampagyruss; Brain 24 to 11: functional connectivity between left thalamus and left fusiform gyrus; Brain 24 to 16: functional connectivity between left thalamus and right fusiform gyrus; Brain 25 to 6: functional connectivity between right dorsolateral prefrontal cortex and thalamus).

4. Discussion

In the present study of patients with acute sTBI, we demonstrate that it is potentially valuable to use resting-state fMRI to detect alterations in FC within and across different brain networks. We identified a substantial proportion of pairwise FC alterations within the DMN and across the DMN-ECN but a lower proportion within the ECN in patients with traumatic coma. Furthermore, the altered FC pairs could be used for early prediction of the 6-month awakening in patients who present with traumatic coma.

It has been reported that the tendency of individuals’ minds to wander is correlated with activity in the DMN,(Qin et al., 2015) and the primary function served by the DMN is introspectively oriented and self-referential activity19. Generally, the DMN shows higher activity during resting-state behaviors than during cognition-demanding tasks20. Studies in visual and auditory perception have revealed that DMN connectivity is influenced by context and incoming inputs.(Sokoliuk et al., 2021) As such, the integrity of the DMN has been thought to be involved in human consciousness. For this reason, previous studies on patients with acute sTBI revealed that both intranetwork DMN correlations and internetwork DMN anticorrelations were necessary for consciousness recovery5. To our knowledge, consciousness has been proven to be associated with thalamostriatal and thalamocortical connections at the functional circuit level8,(Threlkeld et al., 2018, Vanhaudenhuyse et al., 2011). In this study, we expand the structural window of the DMN, and we investigate the DMN “nodes” in both the cortical and subcortical regions, including the cortex, cingulate, thalamus, and cerebellum. In contrast with previous investigations24,(Wilson et al., 1998), we found that intra-DMN FC can be increased or decreased in traumatic coma, suggesting that there was not just a single anticorrelation relationship between intra-DMN FC and coma pathogenesis.

However, DMN connectivity alone is not sufficient for interpreting consciousness mechanisms, and there is emerging evidence that interactions between multiple networks are important for maintaining or recovering consciousness.(Bodien et al., 2017) As two distinct neuronal networks, the DMN is an “intrinsic” system and has been associated with internal processes (rest and introspection), while the ECN “extrinsic” networks have been linked with processes of external input (task cognition)(Boveroux et al., 2010). It has been suggested that the resting conscious brain undergoes spontaneous and dynamic toggling between networks engaged in self-consciousness and environmental consciousness.(Yeshurun et al., 2021) Therefore, we added the ECN in this study to investigate the differential and interaction mechanism of the two networks in traumatic coma.

The findings showed that there were massive pairwise FC alterations in the ECN and in DMN-ECN interactions. It is interesting that we found that the degree of ECN alteration was not as high as that of DMN alteration in patients with traumatic coma. It is also noteworthy that we did not observe significant pairwise FC differences in the ECN between coma and awake patients. This finding suggests that the effects of ECN were not important in predicting awakening from traumatic coma—or at least that these effects could not be easily observed in the acute phase of sTBI. It is also possible that the subtle changes in ECN activity were not detectible during the investigation. Even so, we found that there were pairwise FC alterations in the DMN-ECN interaction between coma and awake patients, indicating that the ECN may particularly affect the functioning of the DMN in awakening regulation. Our current results raise intriguing new questions about whether the DMN is primarily responsible for modulating human consciousness. This issue is worthy of future investigation.

We must be conservative in interpreting existing findings that decreased pairwise FC between the right superior frontal gyrus and right parahippocampal gyrus located within the DMN showed the highest predictive ability (AUC = 0.827) for awakening in patients with traumatic coma. Given the limited number of coma patients, our results do not necessarily provide evidence for direct causality between pairwise FC alteration and awakening from coma. Nevertheless, our findings highlight the importance of cortical-subcortical FC in consciousness recovery, raising the possibility that long-range functional connection alterations can serve as predictive biomarkers for coma recovery. Moreover, our AUC results of pairwise FC alterations provide some quantitative evidence that pairwise FC alterations with high predictive capability are in the DMN.

Our observation that the DMN is more important than the ECN in coma pathogenesis is in line with the current perspective; that is, the DMN is the classically “default” center for integrating internal and external information. Meanwhile, the DMN representations could be invariant to changes in low-level conscious properties that are not associated with changes in high-level cognition behavior.(Sokoliuk et al., 2021) Specifically, the coma phenotype represents the lowest dimension either in the two-dimensional consciousness mechanism27 or in the global-states theory of consciousness.(Zheng et al., 2022) Therefore, it is reasonable to infer that DMN could be primarily responsible for responding to the appearance and disappearance of low-ranking consciousness, while the ECN could be mainly coupled with the alteration of higher-order consciousness. In other words, the recovery from traumatic coma may be nonlinear. Consciousness recovery may starts from single-dimensional arousal and is dominated by the DMN; this is followed by high-dimensional perception which responds to input from multiple networks. The coordinated role of multiple networks in constituting the different components of consciousness is unknown and should be addressed in future investigations.

5. Limitations

First, we did not conduct a comprehensive behavioral assessment to capture the residual brain function of patients with sTBI in the acute phase because the revised coma recovery scale (CRS-R) is more acceptable than the GCS when diagnosing coma in patients with sTBI. Second, we excluded subjects with motion > 2 mm or 2 degrees of rotation in this study and its potential effect on consciousness needs further clarification. In addition, macroscale fMRI data analysis cannot capture the mesoscale circuit and microscale molecular information of sTBI, and future research integrating multimodal and multidimensional data can further clarify this compensation mechanism. Furthermore, it was difficult to select a sample with a similar number of good and poor outcomes because patients with sTBIs have a high probability of mortality and poor neurologic outcomes.

6. Conclusions

In this moderately innovative study, we presented the respective and interactive mechanisms of the DMN and ECN that are related to coma pathogenesis and awakening prediction. We also found the role of the DMN and ECN interaction in modulating traumatic coma. More significantly, we showed that the DMN plays a more important role than the ECN not only in coma pathogenesis but also in awakening prediction. These findings suggested that consciousness may not be regulated by a single functional network but by the coordinated activities of multiple brain networks.

Ethical Approval/Informed Consent

This study was approved by Shanghai Ninth People’s Hospital. All procedures performed were in accordance with the ethical standards of the institutional committee.

Funding

This project is supported by Shanghai Science Committee (No.19411968200).

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.nicl.2023.103361.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Supplementary data 1
mmc1.doc (60KB, doc)
Supplementary data 2
mmc2.doc (52.5KB, doc)

Data availability

The authors do not have permission to share data.

References

  1. Bayne T., Hohwy J., Owen A.M. Are There Levels of Consciousness? Trends Cogn Sci. 2016;20(6):405–413. doi: 10.1016/j.tics.2016.03.009. [DOI] [PubMed] [Google Scholar]
  2. Bianciardi M., Izzy S., Rosen B.R., Wald L.L., Edlow B.L. Location of Subcortical Microbleeds and Recovery of Consciousness After Severe Traumatic Brain Injury. Neurology. 2021;97(2):e113–e123. doi: 10.1212/WNL.0000000000012192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bodien Y.G., Chatelle C., Edlow B.L. Functional Networks in Disorders of Consciousness. Semin Neurol. 2017;37(5):485–502. doi: 10.1055/s-0037-1607310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Boveroux P., Vanhaudenhuyse A., Bruno M.-A., Noirhomme Q., Lauwick S., Luxen A., Degueldre C., Plenevaux A., Schnakers C., Phillips C., Brichant J.-F., Bonhomme V., Maquet P., Greicius M.D., Laureys S., Boly M. Breakdown of within- and between-network resting state functional magnetic resonance imaging connectivity during propofol-induced loss of consciousness. Anesthesiology. 2010;113(5):1038–1053. doi: 10.1097/ALN.0b013e3181f697f5. [DOI] [PubMed] [Google Scholar]
  5. Crone J.S., Ladurner G., Höller Y., Golaszewski S., Trinka E., Kronbichler M., Jiang T. Deactivation of the default mode network as a marker of impaired consciousness: an fMRI study. PLoS One. 2011;6(10):e26373. doi: 10.1371/journal.pone.0026373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Davis M.H., Coleman M.R., Absalom A.R., Rodd J.M., Johnsrude I.S., Matta B.F., Owen A.M., Menon D.K. Dissociating speech perception and comprehension at reduced levels of awareness. Proc Natl Acad Sci U S A. 2007;104(41):16032–16037. doi: 10.1073/pnas.0701309104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Edlow B.L., Chatelle C., Spencer C.A., Chu C.J., Bodien Y.G., O’Connor K.L., Hirschberg R.E., Hochberg L.R., Giacino J.T., Rosenthal E.S., Wu O. Early detection of consciousness in patients with acute severe traumatic brain injury. Brain. 2017;140(9):2399–2414. doi: 10.1093/brain/awx176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Edlow B.L., Claassen J., Schiff N.D., Greer D.M. Recovery from disorders of consciousness: mechanisms, prognosis and emerging therapies. Nat Rev Neurol. 2021;17(3):135–156. doi: 10.1038/s41582-020-00428-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Fransson P. Spontaneous low-frequency BOLD signal fluctuations: an fMRI investigation of the resting-state default mode of brain function hypothesis. Hum Brain Mapp. 2005;26(1):15–29. doi: 10.1002/hbm.20113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Fridman E.A., Beattie B.J., Broft A., Laureys S., Schiff N.D. Regional cerebral metabolic patterns demonstrate the role of anterior forebrain mesocircuit dysfunction in the severely injured brain. Proc Natl Acad Sci U S A. 2014;111(17):6473–6478. doi: 10.1073/pnas.1320969111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. He J.H., Cui Y., Song M., Yang Y., Dang Y.Y., Jiang T.Z., Xu R.X. Decreased functional connectivity between the mediodorsal thalamus and default mode network in patients with disorders of consciousness. Acta Neurol Scand. 2015;131(3):145–151. doi: 10.1111/ane.12299. [DOI] [PubMed] [Google Scholar]
  12. Izzy S., Compton R., Carandang R., Hall W., Muehlschlegel S. Self-fulfilling prophecies through withdrawal of care: do they exist in traumatic brain injury, too? Neurocrit Care. 2013;19(3):347–363. doi: 10.1007/s12028-013-9925-z. [DOI] [PubMed] [Google Scholar]
  13. Leblanc G., Boutin A., Shemilt M., Lauzier F., Moore L., Potvin V., Zarychanski R., Archambault P., Lamontagne F., Léger C., Turgeon A.F. Incidence and impact of withdrawal of life-sustaining therapies in clinical trials of severe traumatic brain injury: A systematic review. Clin Trials. 2018;15(4):398–412. doi: 10.1177/1740774518771233. [DOI] [PubMed] [Google Scholar]
  14. Mason M.F., Norton M.I., Van Horn J.D., Wegner D.M., Grafton S.T., Macrae C.N. Wandering minds: the default network and stimulus-independent thought. Science. 2007;315(5810):393–395. doi: 10.1126/science.1131295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Qin P., Northoff G. How is our self related to midline regions and the default-mode network? Neuroimage. 2011;57(3):1221–1233. doi: 10.1016/j.neuroimage.2011.05.028. [DOI] [PubMed] [Google Scholar]
  16. Qin P., Wu X., Huang Z., Duncan N.W., Tang W., Wolff A., Hu J., Gao L., Jin Y.i., Wu X., Zhang J., Lu L.u., Wu C., Qu X., Mao Y., Weng X., Zhang J., Northoff G. How are different neural networks related to consciousness? Ann Neurol. 2015;78(4):594–605. doi: 10.1002/ana.24479. [DOI] [PubMed] [Google Scholar]
  17. Sokoliuk R., Degano G., Banellis L., Melloni L., Hayton T., Sturman S., Veenith T., Yakoub K.M., Belli A., Noppeney U., Cruse D. Covert Speech Comprehension Predicts Recovery From Acute Unresponsive States. Ann Neurol. 2021;89(4):646–656. doi: 10.1002/ana.25995. [DOI] [PubMed] [Google Scholar]
  18. Threlkeld Z.D., Bodien Y.G., Rosenthal E.S., et al. Functional networks reemerge during recovery of consciousness after acute severe traumatic brain injury. Cortex. 2018;106:299–308. doi: 10.1016/j.cortex.2018.05.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Vanhaudenhuyse A., Demertzi A., Schabus M., Noirhomme Q., Bredart S., Boly M., Phillips C., Soddu A., Luxen A., Moonen G., Laureys S. Two distinct neuronal networks mediate the awareness of environment and of self. J Cogn Neurosci. 2011;23(3):570–578. doi: 10.1162/jocn.2010.21488. [DOI] [PubMed] [Google Scholar]
  20. Wilson J.T., Pettigrew L.E., Teasdale G.M. Structured interviews for the Glasgow Outcome Scale and the extended Glasgow Outcome Scale: guidelines for their use. J Neurotrauma. 1998;15(8):573–585. doi: 10.1089/neu.1998.15.573. [DOI] [PubMed] [Google Scholar]
  21. Yeshurun Y., Nguyen M., Hasson U. The default mode network: where the idiosyncratic self meets the shared social world. Nat Rev Neurosci. 2021;22(3):181–192. doi: 10.1038/s41583-020-00420-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Zheng R., Zhuang Z., Zhao C., Zhao Z., Yang X., Zhou Y., Pan S., Chen K., Li K., Huang Q., Wang Y., Ma Y. Chinese Admission Warning Strategy for Predicting the Hospital Discharge Outcome in Patients with Traumatic Brain Injury. J Clin Med. 2022;11(4):974. doi: 10.3390/jcm11040974. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary data 1
mmc1.doc (60KB, doc)
Supplementary data 2
mmc2.doc (52.5KB, doc)

Data Availability Statement

The authors do not have permission to share data.

Articles from NeuroImage : Clinical are provided here courtesy of Elsevier

RESOURCES