Introduction

Traumatic brain injury (TBI) is a common type of trauma, affecting an estimated 69 million patients annually, and is a significant cause of death and disability worldwide [1]. Patients with TBI frequently experience persistent symptoms including slow reaction time, decreased attention and learning ability, executive dysfunction, and impaired emotional regulation, which impose a huge medical and social burden [2]. Diffuse axonal injury (DAI) is a specific pathological type of TBI that can significantly affect the functional connectivity (FC) of neural circuits [3,4], particularly due to the disconnections of vital network hubs [5], and is an important cause of cognitive, behavioral, and affective deficits in patients with TBI. Cognitive impairment is associated with aberrant structural integrity [6,7] and FC [8,9]in the cortico-subcortical connections. The striatum is organized into subregions, including the caudate, putamen, and nucleus accumbens, as part of the cortico-striato-thalamo-cortical loops, primarily integrating motivational, emotional, learning, cognitive, stress, and sensorimotor information [10]. Previous studies have shown that these discrete functions can be mapped to distinct functional striatal zones [11]. The striatum may be particularly vulnerable to axonal injury because it receives the most cortical projections to the basal ganglia and combines cortical and subcortical information [10]. Several previous studies have suggested that the integrity of the striatum and corticostriatal connectivity is affected following TBI, resulting in executive dysfunction and impaired cognitive flexibility [8,12,13], particularly given the effects of DAI and the susceptibility of dopamine to influence striatal function after TBI [14]. However, whole-brain FC of the striatal subdivision in patients with DAI has received little attention. Studying striatal FC can help elucidate the relationship between striatal networks and cognitive impairment in patients with DAI. This may provide insight into the neural basis of cognitive deficits post-TBI and guide cognitive rehabilitation strategies.

Most previous neuroimaging studies have focused on cortical-cortical interactions; however, altered cortical-subcortical connectivity may also be the key to headache [15], emotion [16], and cognitive deficits [17] in patients after TBI. The striatum plays a crucial role in various motor-related functions, cognitive processes, and emotional regulation [18] as the center of the mesocorticolimbic pathway and receives innervation from multiple cortical regions and the midbrain [19]. It is increasingly clear that the striatum is highly heterogeneous. The inferior (VSi) and superior (VSs) regions of the ventral striatum are primarily involved in the affective limbic system, receiving projections mainly from the prefrontal cortex, anterior cingulate cortex, temporal lobe, and limbic structures [20]. In contrast, the dorsal-caudal putamen (DCP) and dorsal-rostral putamen (DRP), which mainly receive projections from the precentral/posterior central gyrus, play crucial roles in the sensorimotor system [21]. The ventral-rostral putamen (VRP) and dorsal caudate (DC) primarily receive projections from the association cortex and are involved in executive functions [20,22,23]. Disruption of the corticostriatal loops has been observed in various diseases. For instance, reduced bilateral volumes and abnormal FC of the nucleus accumbens have been observed in patients with depression related to late insomnia, lassitude [24], and anhedonia [25,26]. Patients with prodromal Huntington’s disease exhibit disrupted functional and structural connectivity between the cortical regions and caudate, accompanied by impaired motor and executive functions [27,28]. Furthermore, caudate connectivity is associated with executive function in attention deficit hyperactivity disorder [29]. These findings imply that altered corticostriatal connectivity may contribute to cognitive dysfunction following DAI, particularly in relation to executive dysfunction.

Advanced resting-state MRI techniques have provided an important means of studying the effects of TBI on the striatum, which has often been treated as a unified entity or several homogeneous compartments in recent neuroimaging studies. However, the circuitry of the striatum is topographically organized and exhibits more intricate relationships with cortical networks than previously understood. Studies have found that structural and functional changes may occur in the striatum of patients with TBI. For example, decreased executive function in patients with severe TBI is related to damage to the integrity of the ventral striatum and associated structures [12]. A previous structural study that employed graph theory analysis demonstrated that betweenness and eigenvector centralities were reduced within network hubs, including the caudate, which could accurately predict executive dysfunction [5]. Another study showed that the restoration of abnormal shape atrophy in the right dorsal medial caudate is critical for improving cognitive function in patients with mild TBI [6]. Previous studies have found that abnormal caudate volumes in patients with TBI significantly disrupt information processing speed [30], task-switching ability [31], and coma level [32]. Moreover, functional MRI studies have indicated that abnormal caudate activation is associated with impaired cognitive flexibility [11], cognitive fatigue [33], and working memory [34] in patients with TBI. A longitudinal study showed that changes in caudate-based dys FC, mainly distributed in the executive and emotional processing networks, could serve as a neuroimaging biomarker in patients with mild TBI. These studies suggest that the caudate may play a critical role in the cognitive function of patients with DAI. However, the alteration pattern of the FC of striatal subregions remains unclear, particularly considering the effects of DAI on long-distance white matter tracts that connect brain networks [3].

In this study, we directly tested the hypothesis that whole-brain striatal FC was abnormally decreased following DAI using seed-based FC analysis, and that disruption in striatum-based FC correlated with impairments in cognitive function, which may provide a new basis for the neuropathophysiological mechanism of DAI.

Methods Study subjects

A total of 182 patients with TBI were recruited from the Department of Neurosurgery of the First Affiliated Hospital of Nanchang University between April 2013 and December 2017, of whom 26 patients with DAI were retrospectively selected. The inclusion criteria for patients with DAI were as follows: (1) history of typical closed craniocerebral injury (high-speed rotation or rapid acceleration-deceleration); (2) hemodynamic stability to ensure clinical safety and examination cooperation; and (3) age 18–60 years and ultimately diagnosed with DAI characterized by microhemorrhagic foci on the magnetic susceptibility sequence. Exclusion criteria were: (1) prior history of TBI, neurological disorders, or cerebrovascular disease; (2) significant cerebral hemorrhage, cerebral contusion, or subdural or epidural hematoma (volume >10 ml) on computed tomography (CT) and MRI; (4) abnormal findings on MRI (e.g. tumors, epilepsy, and vascular malformations) or the presence of prior neurological or psychiatric disorders; and (5) contraindications to MRI examination. Twenty-seven healthy volunteers were recruited from the community as a control group during the same period. All healthy controls (HCs) had no history of neurological or psychiatric disorders, TBI, or other brain diseases.

The severity of cognitive and executive dysfunction and mood changes in patients with DAI was assessed using appropriate clinical scales prior to MRI scanning examination, including the Glasgow Coma Score (GCS), Motor Function Assessment Scale (MAS), Disability Rating Scale (DRS), Mini-Mental State Examination (MMSE), Activities of Daily Living Scale (ADL), Agitated Behavior Scale (ABS), Hamilton Anxiety Scale (HAMA), Clinical Dementia Rating Scale (CDR). Additionally, all subjects were right-handed. The study followed the Declaration of Helsinki and was approved by the Medical Ethics Committee of the First Affiliated Hospital of Nanchang University (Approval No. 2015030). All subjects and/or their family members provided informed consent and signed informed consent forms before the examination.

Magnetic resonance data acquisition

Resting-state functional MRI data acquisition was performed using a German Siemens Trio Tim 3T superconducting MR scanner with an 8-channel phased-array head orthogonal coil by employing a scanning GRE-EPI sequence with the specific parameters of TR 2000 ms, TE 30 ms, FOV 200 × 200; matrix, 64 × 64; flip angle, 90°; layer thickness, 4 mm; and continuous scanning of 30 layers. The whole process took 8 min, and a total of 240 time phases. During the scanning process, patients were instructed to remain quiet, not think, breathe calmly, close their eyes, and remain awake. 3D-T1WI sagittal scanning was performed using the following specific parameters: repetition TR 1900 ms, TE 2.26 ms, voxel size, 1 mm × 1 mm × 1 mm; matrix, 256 × 256; and a total of 176 layers were acquired. All subjects were scanned using a magnetic susceptibility-weighted imaging sequence with TR 28 ms, TE 20 ms; FOV = 240 × 240, matrix 448 × 384, flip angle 15°, bandwidth 15.6 kHz, and acquisition layer thickness of 1.2 mm. CT and conventional MRI plain scanning were also performed to exclude brain diseases.

Data preprocessing

For data preparation in this study, Data Processing and Analysis of Brain Imaging (v6.0 http://www.rfmri.org/dpabi were based on the Statistical Parametric Mapping (SPM12, https://www.fil.ion.ucl.ac.uk/spm/software/spm12/) toolkits on the MATLAB 2016b platform (The MathWorks, Inc., Natick, MA, USA). The specific steps were as follows: the first 10 volumes of resting-state fMRI data were discarded, and the remaining 230 volumes acquired from each participant were corrected for differences in slice acquisition times. The resulting images were then realigned to correct for small movements between scans. Images with head motion that exceeded 2-mm translation or 2° rotation during functional imaging were excluded. Individual T1-weighted structural images were co-registered with mean-realigned echo-planar images. The transformed structural images were then segmented into gray matter, white matter, and cerebrospinal fluid, and registered in the Montreal Neurological Institute space using the Diffeomorphic Anatomical Registration Through Exponentiated Lie Algebra toolbox. The same transformation parameters were applied to the functional images for spatial normalization, and the volumes were resliced to a resolution of 3 mm × 3 mm × 3 mm. Spatial smoothing with a full-width half-maximum 6-mm Gaussian kernel was used to improve the signal-to-noise ratio. A nuisance linear regression was performed to reduce the effects of confounding factors using 24 head-motion parameters (Friston 24-parameter model), white matter, cerebrospinal fluid, and whole-brain signal as covariates. Finally, temporal filtering (0.01–0.1 Hz) of the time series was performed.

Functional connectivity analysis

This study implemented seed-based FC analysis, as in previous studies [18,20,22,35]. Bilateral spherical regions of interest (ROIs), with a radius of 6 mm were placed within six predefined subdivisions of the striatum, representing the affective limbic, sensorimotor, and executive loops. The ROIs for the affective limbic loop were the VSi at the coordinates (±9, 9, −8) and the VSs at coordinates (±10, 15, 0). The ROIs for the sensorimotor loop were the DCP at coordinates (±28, 1, 3) and the DRP at coordinates (±25, 8, 6). Finally, the ROIs for the executive loop were the DC at coordinates (±13, 15, 9) and the VRP at coordinates (±20, 12, ± 3) (Table 1). The average time series of the seed regions was extracted using the seed-point correlation analysis method, where a sphere with a radius of 6 mm was employed. Subsequently, Pearson’s correlation analysis was conducted to calculate the correlation coefficients between the time series of the seed regions and the time series of every voxel across the whole brain. These correlation coefficients represented the FC between the seed regions and other voxels in the brain. To enhance normality, individual correlation matrices were transformed into a z-score matrix using Fisher r-to-z transformation. This resulted in a brain map illustrating the FC of the 12 subregions within the striatum for each subject.

Table 1 - Distribution of 12 seeds in striatal subregions Seed abbreviations MNI coordinates (mm) x y z Inferior ventral striatum VSi ±9 9 −8 Superior ventral striatum VSs ±10 15 0 Dorsal caudal putamen DCP ±28 1 3 Dorsal rostral putamen DRP ±25 8 6 Dorsal caudate DC ±13 15 −9 Ventral rostral putamen VRP ±20 12 −3

−, left; +, right; MNI, Montreal Neuroscience Institute.

Statistical analysis

Group-level independent two-sample t-tests were conducted to compare the differences in FC based on each striatal seed between individuals with DAI and HCs. Multiple comparison corrections were performed using the Gaussian Random Field (GRF) theory (at the level of voxels P < 0.001, level of clumps P < 0.05, GRF corrected). Pearson’s correlation coefficients between the mean FC values of remarkably different brain clumps and each clinical behavior scale of patients with DAI were further computed. These correlations were controlled for age, sex, and head-motion parameters (mean FD). The measurement data were analyzed using IBM SPSS 26 software. Comparisons of sex between the two groups of subjects was performed by chi-square test, and two independent samples t-tests were used to analyze the information of age, education, head movement parameters, and clinical scales, and the differences were deemed to be statistically significant at P < 0.05.

Results Demographics

There were no statistically significant differences in age, sex, or educational level between the 2 groups. As shown in Table 2, the clinical scale scores of the DAI group were significantly (P < 0.05) lower than those of the control group for the GCS, DRS, MRS, HAMA, CDR, MMSE, and ADL.

Table 2 - Comparison of clinical characteristics among the DAI and HC groups Characteristics DAI(n = 26) HC (n = 27) P-value Sex (male/female) 18/8 20/7 0.64a Age (years) 38.35 ± 14.59 38.26 ± 13.32 0.98b Education (years) 7.07 ± 3.49 7.11 ± 2.95 0.97b Injury-to-MRI interval (days) (median, range) 24(2–210) — — Handedness (right), % 100 100 1.0b GCS (scanning) 13.73 ± 2.38 15 ± 0 0.01b MMSE 19.50 ± 10.48 29.78 ± 0.42 <0.01b DRS 9.31 ± 6.20 0 ± 0 <0.01b MAS 35.88 ± 11.61 47.93 ± 0.27 <0.01b ABS 32.31 ± 12.60 14 ± 0 <0.01b CDR 1.21 ± 0.97 0 ± 0 <0.01b ADL 33.38 ± 14.54 14.33 ± 0.62 <0.01b HAMA 10.62 ± 6.89 1.52 ± 1.01 <0.01b

-, no data.

ABS, Agitated Behavior Scale; ADL, Activities of Daily Living Scale; CDR, Clinical Dementia Rating Scale; DAI, diffuse axonal injury; DRS, Disability Rating Scale; GCS, Glasgow Coma Score; HAMA, Hamilton Anxiety Scale; HC, healthy control group; M, median; MAS, Motor Assessment Scale; MMSE, Mini-Mental State Examination Scale.

aChi-square test.

bTwo-sample t-test.

Between-group results of functional connectivity

Compared to HCs, the DAI group demonstrated significantly decreased FC between the right VSi and right inferior frontal gyrus, as well as right inferior parietal lobule, left VSi and right inferior frontal gyrus, right VSs and bilateral cerebellar posterior lobe, bilateral DCP and right anterior cingulate gyrus, and right DCP and right inferior parietal lobule. Moreover, decreased FC was found between the left DC and right cerebellar posterior lobe, whereas increased FC was found between the left DC and right inferior parietal lobule (Figs. 1 and 2 and Table 3; P < 0.05, GRF correction). There were no significant differences between the two groups in FC in the brain regions of the bilateral VRP, bilateral DRP, left VSs, left DCP, or right DC.

Table 3 - Distribution of brain regions with different functional connectivity between groups ROI Region MNI coordinates Number of voxels t-value P-value x y z R-VSi R-inferior frontal gyrus 3 6 −9 153 −5.219 P < 0.05 R-VSi R-inferior parietal lobule 60 −51 45 134 −4.838 P < 0.05 L-VSi R-inferior frontal gyrus 21 24 −27 197 −5.874 P < 0.05 R-VSs L-cerebelum posterior lobe −24 −81 −57 87 −5.122 P < 0.05 R-VSs R-cerebelum posterior lobe 38 −81 −51 96 −5.448 P < 0.05 R-DCP R-anterior cingulate gyrus 9 6 27 73 −6.281 P < 0.05 R-DCP R-inferior parietal lobule 54 −57 48 50 −5.631 P < 0.05 L-DCP R-anterior cingulate gyrus 9 6 27 124 −6.226 P < 0.05 L-DC R-cerebelum posterior lobe 12 −84 −39 143 −5.351 P < 0.05 L-DC R-inferior parietal lobule 36 −51 60 63 4.363 P < 0.05

T-value is negative for weakened connectivity; t-tests for two-sample, P < 0.001 at the level of voxels, P < 0.05 at the level of clumps, and GRF correction.

DC, dorsal caudate nucleus; DCP, dorsal cribriform nucleus caudalis; L, left; MNI, Montreal Neuroscience Institute; R, right; ROI, region of interest; VSi, inferior ventral striatum; VSs, superior ventral striatum

Fig. 1:

Distribution of brain regions with differences in functional connectivity between groups. Compared with the healthy control group, warm colors represent significantly higher FC values in the DAI group, and the cool colors represent significantly reduced FC values in the DAI group. DAI, diffuse axonal injury; DC, dorsal caudate; DCP, dorsal caudal putamen; HC, healthy control; L, left; R, right; VSi, ventral striatum inferior; VSs, ventral striatum superior.

Fig. 2:

Overlapping distribution of brain regions with differences in functional connectivity between groups. Compared with the healthy control group, warm colors represent significantly higher FC values in the DAI group, and the cool colors represent significantly reduced FC values in the DAI group. DAI, diffuse axonal injury; HC, healthy control; L, left; R, right.

Correlation analysis of brain regions with abnormal FC values and clinical variables in the DAI group

Brain regions with statistically significant differences in FC values between the DAI and HC groups showed a positive correlation between the right anterior cingulate gyrus and GCS score, a negative correlation between the right inferior parietal lobule and DRS score, and a positive correlation between the right anterior cingulate gyrus and MAS score, as well as ABS score. The left cerebellar posterior lobe was negatively correlated with DRS, HAMA, CDR, and ADL scores, and positively correlated with MAS and ABS scores. The right posterior cerebellar lobe was negatively correlated with the DRS score and positively correlated with the MAS, ABS, and MMSE scores (P < 0.05, uncorrected; Table 4). In addition, there was no significant correlation between the FC values of other significantly altered brain regions and the clinical parameters in the DAI group (P > 0.05).

Table 4 - Correlations between altered voxel-based functional connectivity and clinical variables ROI Region (FC values) Correlation coefficient (P value) GCS DRS MAS ABS HAMA CDR MMSE ADL R-DCP R-anterior cingulate gyrus 0.455
(0.020)* −0.308
(0.126) 0.094
(0.647) 0.177
(0.387) 0.076
(0.714) −0.269
(0.184) 0.205
(0.316) −0.204
(0.317) R-DCP R-inferior parietal lobule 0.064
(0.756) −0.448
(0.022) 0.455
(0.020)* 0.443
(0.023)* −0.234
(0.250) −0.032
(0.878) −0.070
(0.735) −0.178
(0.384) R-VSs R-cerebelum posterior lobe 0.201
(0.325) −0.400
(0.043)* 0.484
(0.012)* 0.561
(0.003)** −0.378
(0.057) −0.352
(0.078) 0.394
(0.046)* −0.33
(0.100) R-VSs L-cerebelum posterior lobe 0.159
(0.439) −0.526
(0.006)** 0.520
(0.007)** 0.391
(0.048)* −0.397
(0.045)* −0.448
(0.022)* 0.356
(0.074) −0.481
(0.013)*

ABS, Agitated Behavior Scale; ADL, Activities of Daily Living Scale; CDR, Clinical Dementia Rating Scale; DCP, dorsal cribriform nucleus caudalis; DRS, Disability Rating Scale; FC, functional connectivity; GCS, Glasgow Coma Score; HAMA, Hamilton Anxiety Scale; MAS, Motor Assessment Scale; MMSE, Mini-Mental State Examination Scale; R, right; VSs, superior ventral striatum.

*P < 0.05;

**P < 0.01 uncorrected.

Discussion

The present study explored whole-brain FC changes in striatal subregions of patients with DAI and HCs using a seed-based approach. The results showed that the FC values of the striatum in the DAI group were reduced in the right inferior frontal gyrus, right inferior parietal lobule, anterior cingulate gyrus, and bilateral cerebellar posterior lobe compared to those in the HC group, suggesting that there may be abnormal patterns of corticostriatal and striatum-cerebellum FC in patients with DAI, mainly in the cognitive control, executive, and emotion-processing networks. In addition, we found that the FC values of the left DC nucleus and right inferior parietal lobule increased, indicating that compensatory mechanisms may exist. These results suggest that the striatal subregions may play a critical role in the cognitive function of patients with DAI.

The VSi mainly receives projections from the orbitofrontal cortex, medial prefrontal cortex, anterior cingulate gyrus, temporal lobe, and limbic structures and is inextricably linked to emotional/motivational activities in the brain [20,22]. Our study found decreased FC values in the bilateral VSi and right inferior frontal gyrus in patients with DAI, indicating a disruption of striatal-frontal cortex interactions. Previous studies have demonstrated that the inferior frontal gyrus is involved in the regulation of affective states and cognitive functions, which may be associated with emotional recognition and social communication deficits in patients with TBI [36,37]. Morphological alterations in cortical thickness of the bilateral prefrontal cortex were found in patients with mild TBI, and was associated with cognitive disorder [38]. Some studies have reported a decreased basal metabolic rate and diminished activity in the inferior frontal gyrus of patients with TBI, which may be attributed to their impaired consciousness [39]. A structural study revealed that frontal-subcortical projections were impaired in patients with TBI, especially in the caudate nucleus, and were correlated with task-switching deficits [40]. Our study suggests that the compromised inferior ventral striatal-frontal cortical connections may be related to the potential disruption of structural connectivity induced by DAI, further contributing to cognitive dysfunction.

The inferior parietal lobule is one of the higher contact cortices of the brain, which is responsible for receiving and integrating information from different modalities of somatomotor, auditory, and visual functions, and plays an important role in higher cognitive functions such as language and tool use. Bonnelle et al. [41] found that the bilateral inferior parietal lobules of patients with TBI showed decreased whole-brain FC associated with sustained attention. Consistent with these results, the FC of the right VSi and DCP with the right inferior parietal lobule in the DAI group was diminished, implying a decline in nerve conduction between brain regions in the present study. Furthermore, the FC values of the right inferior parietal lobule were negatively correlated with DRS scores and positively correlated with MAS and ABS scores, suggesting that they may be related to executive control and motor dysfunction in patients with DAI. Meanwhile, we also discovered that the FC values of the left DC and the right inferior parietal lobule were increased, which may be that in order to ensure that the brain function is up and running, the organism further stimulates the caudate nucleus for the purpose of functional compensation and is related to the mechanism of organismal compensation caused by cognitive dysfunction. The caudate nucleus plays a crucial role in miscellaneous cognitive functions [42], and patients with mild TBI exhibit diminished caudate-subcortical FC during the acute phase, which is mainly distributed in the executive and emotional processing networks [43]. The DC nucleus is a key component of the dorsal striatum and is anatomically and functionally connected to high-level cortical areas (e.g. the dorsal or ventral lateral prefrontal cortex, anterior cingulate cortex, and subparietal lobule), and their functional pathways facilitate cognitive control [22]. Corticostriatal structural and functional pathways are connected to the prefrontal and anterior cingulate cortices via the striatum into the basal ganglia region and are associated with executive control processes. The anterior cingulate gyrus is vital for cognitive control and is involved in several processes, including reward processing, conflict monitoring, and action selection [44]. Abnormal changes in corticostriatal connectivity are essential for executive dysfunction after TBI. Some findings have pointed out that the disruption of connectivity between the anterior cingulate gyrus and the right caudate nucleus provokes impairments in executive cognitive control, inducing behavioral disorders and disabilities [8]. Our study found decreased FC between the bilateral VSi and the right inferior frontal gyrus, right DCP, and anterior cingulate gyrus loops. Additionally, the FC values of the right DCP and anterior cingulate gyrus were positively correlated with GCS scores, indicating a potential connection between the severity of clinical symptoms and executive dysfunction in patients with DAI.

Furthermore, impaired FC of the striatal-cerebellar loop was observed in this study, manifested by reduced FC values between the right VSs and bilateral posterior cerebellar lobes, and the left DC and right posterior cerebellar lobes. Further correlation analyses revealed that the FC values of the striatal-posterior cerebellum were correlated with several clinical scales, including cognitive function, locomotion, executive control, and affective anxiety. These results suggest that disruption of the FC of the striatal-cerebellar loop is strongly correlated with the prognosis of patients and can be used as a neuroimaging biomarker to predict the clinical symptoms associated with post-injury in patients with DAI. The cerebellum plays a role in sensorimotor and vestibular control, as well as cognitive, emotional, and autonomic functions. The posterior cerebellar lobe is involved in various neural activities, including motor, emotional, and higher cognitive processing [45]. It is an integral part of the cognitive function process, and aberrant changes in the posterior cerebellar lobe may lead to cerebellar cognitive-emotional syndromes, including deficits in executive function and visuospatial processes, and impairments in language ability and regulation of emotions. Previous studies have reported lower cerebellar metabolism [46] and diminished activity within the cerebellar hemispheres in patients with TBI [47]. Our previous studies also observed reduced FC between the bilateral posterior cerebellar lobes and increased temporal variability of the dynamic fractional amplitude of low-frequency fluctuation in the left posterior cerebellar lobe and bilateral caudate nucleus [48]. It has been proposed that the integrity of the connectivity pathway between the cerebellum and cerebral hemispheres may be disrupted in patients with DAI, which leads to impaired FC , giving rise to relevant clinical symptoms.

The current study has several limitations. First, the sample size of patients with DAI is small owing to the heterogeneity of patients with TBI, and considerably larger samples are needed to further validate our findings. Second, the conditions of patients with DAI are comparatively stable, ensuring clinical safety, and the degree and duration of injuries vary among patients with DAI. Third, the present study investigated the alteration patterns of striatal subdivision resting-state FC, but the resting-state condition may be not a true resting-state condition with noisy scanner sounds, physiological noise and psychologically demanding, which resulted in the influence on brain activity in response to regular stimulation for several people [49,50]. Meanwhile, The human brain is inherently dynamic, and FC in the resting state fluctuates over time. These time-varying states contain a wealth of time-effect information. We will investigate the shared and specific patterns of dynamic FC variability of striato-cortical circuitry in DAI and HC in next work. Fourth, the fine delineation of the striatum may affect the result of FC. Therefore, our further study will focus on the fined parcellation of the striatum subregions. Finally, We only examined resting-state fMRI data in this study, the structural alterations in the striatum were unclear. A longitudinal analysis in conjunction with functional and structural MRI data is needed to observe the pattern of evolution over the course of the disease in further study.

Conclusion

This study provides evidence for abnormalities in cortico-striatal and cerebellar-striatal FC in patients with DAI using seed-based analysis, which is associated with post-injury cognitive and motor disorders, and providing potential neuroimaging markers for the diagnosis and treatment of DAI.

Acknowledgements

This study was supported by the Jiangxi Province Natural Science Foundation (20232BAB216044), the Science and Technology Plan of Jiangxi Provincial Health Commission (202212861), and Postgraduate Innovation Special Funding Project (YC2023-B066).

Conflicts of interest