**1. Introduction**

Mild traumatic brain injury (mTBI) is an important but less recognized public health concern [1], which accounts for nearly 80% of all traumatic brain injuries [2,3]. Patients with mTBI have impaired executive function, which disrupts the normal performance of daily activities [4]. Executive function is a high-level cognitive function in human beings [5]. As a core component of the executive function, cognitive flexibility defines the ability of individuals to constantly adjust their behavioral responses according to the changing external environment [6,7]. Many studies have demonstrated executive dysfunction in patients with various diseases and the patients exhibit impaired cognitive flexibility [8–12].

Previous studies conducting magnetic resonance imaging (MRI) found that patients with traumatic brain injury have executive dysfunction, and showed impaired cognitive flexibility and altered information processing speed on the behavior level [13]. A previous

**Citation:** Xu, H.; Zhang, X.; Bai, G. Abnormal Dorsal Caudate Activation Mediated Impaired Cognitive Flexibility in Mild Traumatic Brain Injury. *J. Clin. Med.* **2022**, *11*, 2484. https://doi.org/10.3390/ jcm11092484

Academic Editors: Nada Andelic, Cecilie Røe, Eirik Helseth, Emilie Isager Howe, Marit Vindal Forslund and Torgeir Hellstrom

Received: 17 March 2022 Accepted: 26 April 2022 Published: 28 April 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

study that employed graph theory analysis showed that patients exhibited reduced centrality of characteristic vectors of caudate and cingulate cortex, which could accurately predict executive dysfunction in traumatic brain injury [13]. Another study that conducted a wordbased working memory task in patients with severe traumatic brain injury showed that the abnormal activation response of caudate was negatively correlated with the patient's cognitive fatigue under complex conditions, but was positively associated with cognitive fatigue under simple conditions [14]. In addition, one task-switching study showed that patients with local caudate atrophy had abnormal cognitive flexibility, and needed sufficient cognitive load during task switching, but exhibited a significantly increased error rate [15].

Besides, other studies demonstrated that patients with traumatic brain injury had widespread altered white matter microstructure, and the damage to the upper radiation crown from the caudate to the anterior auxiliary motor zone was significantly associated with the switching cost required for patients to achieve task switching, which could predict impaired cognitive flexibility in the patients [16]. Therefore, patients with traumatic brain injury exhibited impaired cognitive flexibility with altered behavioral responses as well as structural and functional abnormalities in the hub region caudate. However, these studies mainly focused on the cognitive dysfunction in patients with moderate to severe traumatic brain injury.

Moreover, previous studies mostly employed specific neuropsychological measurement tools such as the Wisconsin Card Sorting Test to assess cognitive function [17], but were limited by subjective judgments and the expectation effect; it has been proved that when individuals with mild head injury are informed of this, they may experience cognitive difficulties and perform worse on neuropsychological tests compared to the individuals who are uninformed [18,19]. In addition, other studies conducted different psychological experimental tasks to measure cognitive functions in patients with traumatic brain injury, such as working memory tasks [20–22] and the go/no-go task [23,24], which were performed mainly to measure working memory and inhibitory control ability, respectively. Furthermore, few studies have investigated cognitive flexibility in mTBI patients and its associated neural mechanisms using objective tools such as the psychological experiment paradigm.

To explore underlying neural mechanisms of cognitive flexibility in mTBI patients, we performed a rule-based cognitive control experimental paradigm with functional MRI. Here, we first investigated specific behavioral patterns of cognitive impairment in the mTBI patients, and then explored abnormal brain activation in task conditions. Thereafter, we assessed the relationship between altered brain activation, information processing speed and cognitive flexibility. We aimed to identify neural correlates of impaired cognitive flexibility in mTBI.

#### **2. Materials and Methods**

#### *2.1. Participants*

Sixteen mTBI patients (6 females, mean age 25.8 ± 2.8 years) were recruited from the local emergency department. Diagnosis of the mTBI was assessed by two experienced neurologists following the World Health Organization's Collaborating Centre for Neurotrauma Task Force [4]. To be included in this study, all mTBI patients had to have met the following inclusion criteria: (1) a Glasgow Coma Scale score of 13–15; (2) one or more of loss of consciousness (if present) <30 min, post-traumatic amnesia (if present) <24 h, and/or other transient neurological abnormalities such as focal signs, seizure, and intracranial lesion not necessitating surgery. We excluded patients with a history of neurological disease, longstanding psychiatric condition, head injury, substance or alcohol abuse, clinical symptoms of depression and anxiety, intubation and/or presence of a skull fracture as well as administration of sedatives on arrival in the emergency department, spinal cord injury. Patients with a manifestation of mTBI due to medications by other injuries (e.g., systemic injuries, facial injuries, or intubation) or other sources such as psychological trauma, language barrier, or coexisting medical conditions as well as those caused by penetrating craniocerebral

injury were also excluded from this study. In addition, 17 age- and sex-matched healthy controls (HCs) were also enrolled (5 females, mean age 27.8 ± 3.3 years).

#### *2.2. Neuropsychological Tests*

Several neuropsychological performance tests were performed. The tests included Trail-Making Test Part-A (TMT-A) for rote memory assessment, Forward Digit Span (FDS) and Backward Digit Span (BDS) test of Wechsler Adult Intelligence Scale-III for working memory assessment and Digit Symbol Coding (DSC) task for cognitive function assessment and information processing speed. On the other hand, we employed self-reported symptomatology assessments such as Insomnia, which was evaluated using the Insomnia Severity Index (ISI) for sleep quality and short-form Headache Impact Test (HIT) for severity of headaches. All the neuropsychological tests were performed by an experienced clinical psychologist blinded to this study.

#### *2.3. Experimental Design and Procedures*

We employed a modified version of the rule-based task-switching experimental paradigm [25–27], where participants responded to the target digital stimuli based on the cues presented initially. It is a highly time-efficient event-related fMRI paradigm, which has been designed to specifically probe for cognitive flexibility and stability. Participants were instructed to perform one of three task conditions on numerical stimuli based on a cue presented simultaneously during each trial (Figure 1). The task was generated and presented using PsychToolBox and appeared on a uniform black background [28,29]. At each trial, participants were cued explicitly (using a square or diamond cue) as to which condition should be performed during the digit stimuli between 1 and 9 (excluding number 5). Three conditions were set in the experiment: ongoing (OG), distractor inhibition (DI) and task switch (TS). For the ongoing (OG) condition, a diamond cue was presented at the center of a screen with a digit stimulus on the left side, and participants were asked to indicate whether the digit was larger or smaller than five. On the other hand, for the distractor inhibition (DI) condition, a diamond cue was presented at the center of a screen with two-digit stimuli on each side, and participants were asked to indicate whether the left digit was larger or smaller than five, and had to inhibit their response to the right digit (assessing cognitive stability). For the task switch (TS) condition, a square cue was presented at the center of a screen with two-digit stimuli at each side, and the participants were asked to switch from the left digit to the right digit and then decide whether the right digit was odd or even (assessing cognitive flexibility). At each trial, the condition cue and digit stimuli were simultaneously presented for 2000 ms, followed by a variable inter-trial interval of 2000, 4000, or 6000 ms. Participants were required to accurately respond as quickly as possible within the limit of 2000 ms.

All participants received out-of-scanner practice with trial-to-trial feedback and were instructed to provide accurate and quick responses until they attained 95% accuracy. The task was split into two functional scanning runs of 84 trials each. Each run started and ended with two dummy scans (5 s) for scanner signal stabilization, while the participants looked at a fixation cross at the center of the screen.

**Figure 1.** Schematic illustration of the rule-based task-switching experimental paradigm. During three task conditions [ongoing (OG) condition, distractor inhibition (DI) condition, task switch (TS) condition], depending on task cues (square vs. diamond), participants performed one of the different tasks on visually presented number stimuli (smaller/larger than 5 vs. odd/even).

#### *2.4. fMRI Data Acquisition*

Functional MRI images were acquired on a 3.0 Tesla MRI scanner (GE 750 Medical Systems), equipped with a single-shot, gradient-recalled echo planar imaging (EPI) sequence and a 32-channel head coil.

A total of 340 functional volumes were acquired in two runs, using a T2\*-weighted BOLD-sensitive gradient-recalled, EPI sequence with 54 slices covering the whole brain [repetition time (TR) = 2500 ms, echo time (TE) = 30 ms, slice thickness = 3 mm, flip angle (FA) = 90◦, field of view (FOV) = 216 mm × 216 mm, matrix size = 64 × 64, voxel size = 3 mm × 3 mm × 3 mm]. The first and last two volumes of each run were discarded to allow for stable magnetization.

Before performing functional MRI scan, a high-resolution T1-weighted magnetization prepared-rapid gradient echo scan was acquired with the following parameters: TR = 2300 ms, TE = 3.17 ms, FA = 9◦, slice thickness = 1 mm, FOV = 256 mm × 256 mm, matrix size = 256 × 256.

We also captured multiple neuroimaging data (including T1-flair, T2-flair, T2, susceptibility-weighted imaging (SWI)) for all the mTBI patients. These data were used to assess the presence of focal lesions and cerebral microbleeds. Visible contusion lesions were not detected in any of the patients.
