Detection of Abnormal Spontaneous Brain Activity Patterns in Patients With Orbital Fractures Using Fractional Amplitude of Low Frequency Fluctuation

Background To date, no in-depth study has been conducted on the intrinsic pathological relationship between altered brain activity and related behavioral changes in patients with orbital fracture (OF). Purpose The present research aimed to explore the potential functional network cerebrum activities in patients with OF using resting state functional magnetic resonance imaging–fractional amplitude of low-frequency fluctuation (rsfMRI-fALFF). This technique can reveal dynamic functional changes in specific cerebrum areas. Methods Twenty patients with OF and 20 healthy controls (HCs) were included, closely matched in terms of gender, age, weight, and education level. To record spontaneous cerebral activity changes, the rsfMRI-fALFF tool was applied. Receiver operating characteristic (ROC) curves and Pearson's correlation analysis were used to analyze mean fALFF values in specific cerebrum regions and to explore changes of behavioral changes in patients with OF. The Hospital Depression and Anxiety scale was applied to reveal the relationship between emotional states and fALFF values of the right superior temporal gyrus in patients with OF. Results In comparison with HCs, significantly lower fALFF values were detected in the left anterior cingulate gyrus (LACG) and right superior temporal gyrus (RSTG) in patients with OF. ROC curve analysis showed excellent accuracy. The mean fALFF values of the RSTG negatively correlated with the depression score as well as the anxiety score. Conclusion The finding of abnormal spontaneous activities in cerebral regions may contribute to more comprehensive understanding of the potential neural network changes in patients with OF. The changes of fALFF value in patients with OF may help to gauge their emotional changes and clinical recovery levels.


INTRODUCTION
The orbital bones are fragile, with no protective surround, making them vulnerable to orbital wall fracture (1), of which trauma is the main cause. Previous research (2) has shown that about half of orbital fractures are isolated and are usually limited to one orbital wall, the orbital floor, and medial wall being the most frequently occurring fracture sites (3). Orbital fractures mostly occur in children and young people, and are more common in males than females (4,5). Chi et al. reviewed 733 cases of orbital fractures, among which three quarters were male (6). Conservative treatment is often used for small orbital fractures without displacement (7), while, for large displaced fractures, surgical intervention is necessary (8,9). Orbital fractures may cause exophthalmos (10), enophthalmos (11), diplopia (12), entropion (13), subconjunctival hemorrhage (14), and even blindness (15). Therefore, early monitoring and termination of adverse disease progression in patients with OF are very important. The use of modern imaging technology to study the brain activity of patients with OF may be important as a means by which to improve understanding of the mechanism of potential pathological changes in this condition, and may, therefore, be beneficial to the management of complications. Previous studies have confirmed that the changes of spontaneous brain activity in related brain regions can be used as an indicator of disease progression. Therefore, we tried to explore the value of using modern imaging techniques to explore the value of spontaneous brain activity changes as a marker of disease progression in patients with orbital fractures.
Magnetic resonance imaging (MRI), as a widely used auxiliary imaging technology, was developed in the 1980s and provides us a preliminary understanding of the anatomical structure and operating mechanism of the brain (16,17). Hemodynamic changes caused by neuronal activity can be qualitatively measured with the help of MRI technology, known as functional magnetic resonance imaging (fMRI). This approach has been used in a variety of studies on the mechanism and effects of spontaneous neuronal activity in the brain, and has been helpful in exploring the pathophysiological changes and pathogenesis of various diseases (18,19). The fractional amplitude of low frequency fluctuation (fALFF), a resting state fMRI method, has provided an index for the evaluation of spontaneous neural activity, and its accuracy and sensitivity have been widely confirmed (16). To our knowledge, the present experiment was the first attempt to explore the connection between spontaneous brain activity and behavioral performance in patients with OF using the fALFF method as well as to explore the value of fALFF in evaluating the pathological changes and severity of OF.
The 20 HCs (12 males, 8 females) were highly similar to the OF group in sex, age, weight, and education level. Our study met the ethical standards of the Medical Ethics Committee of the First Affiliated Hospital of Nanchang University as well as the principles of the Declaration of Helsinki. After materials, methods, purpose, and underlying risks of this experiment were explained, each participant signed a declaration of informed consents.

MRI Parameters
MRI scanning was conducted using a Trio 3-Tesla MR scanner (Trio; Siemens, Munich, Berlin, Germany) in all the participants. During the MRI scanning, other interference factors were excluded, and the subjects remained awake, breathing normally and with good vital signs. The whole-brainT1-weights were obtained with the application of the spoiled gradient-recalled echo sequence. Relevant corresponding parameter settings of structural images were as follows: echo time = 2.25 ms, repetition time = 1,800 ms, field of view = 250 × 250 mm 2 , layer interval = 0.5 mm, flip angle = 90 • , matrix = 256 × 256, thickness = 1 mm. Functional images (n = 240) were captured with the following settings: echo time = 30 ms, repetition time = 2,000 ms, field of view = 220 × 220 mm 2 , flip angle = 90 • , matrix = 64 × 64, thickness = 4 mm.

FMRI Data Processing
All data were pre-filtered using MRIcro (www.MRIcro.com) and then preprocessed the filtered data using SPM8 (https://www.fil. ion.ucl.ac.uk/spm/). In pre-filtering, the first 10 volumes were  regarded as invalid data and excluded to ensure steady signals. The volumes were offset by no more than 2 mm in X, Y, or Z directions. On the basis of the standard echo planar imaging template, the images were resampled and normalized (with a standard setting of voxel size 3 × 3 × 3 mm, and were smoothed) to enhance the signal-to-noise ratio. This method has been described in detail previously (20).

FALFF Analysis
To calculate fALFF, a full-width Gaussian kernel (half maximum = 6 × 6 × 6 mm 3 ) was used to smooth the remaining 230 images. Band-pass (0.01-0.08 Hz) filtering was used to control for movement artifacts and low frequency drift. A fast Fourier transform (FFT) algorithm was used to obtain the signal power spectrum, and fAFLL was calculated as the ratio of the amplitude at each value in the low frequency band (0.01-0.08 Hz) to fullband (0-0.25 Hz) power amplitude.

Brain-Behavior Correlation Analysis
To look for any associations between brain activity and behavioral performance, brain regions of interest were determined based on fALFF values, and Pearson's correlation analysis was used to explore the linear relationship between activities in these regions and clinical manifestations.

Statistical Analysis
Using SPSS software version 20.0 (IBM Corp, Armonk, NY, USA), an independent sample t-test was conducted on the common clinical variables and demographic data of patients with OF and HCs using a 5% significance level. A two-sample t-test was used to compare the functional data. Based on Gaussian random field theory, the statistical threshold of the voxel level in multiple comparisons was set at p < 0.05. Gaussian random field theory was used to determine the significance of the functional image at the 5% level with a cluster size > 40 voxels.
Using the mean fALFF in various cerebral regions of HCs and patients with OF, the areas under the ROC curves (AUC) were obtained. In addition, Pearson correlation analysis was used to look for associations between the mean fALFF values in multiple cerebrum regions and characteristics of clinical behavior in patients with OF.

Differences in FALFF
In comparison with HCs, the patients with OF showed significant lower fALFF values in the left anterior cingulate gyrus (LACG) and right superior temporal gyrus (RSTG) (Figure 1, Table 2). The mean fAFLL values are shown in Figure 2.

Receiver Operating Characteristic Curves
ROC curves were used to visualize the comparison between average fALFF values of patients with OF and HCs, and the areas under the curves (AUCs) were used as indicators of diagnostic accuracy. Using this approach, AUCs of the LATG and RSTG were found to be 0.983 and 1.000, respectively (Figure 3).

DISCUSSION
To our knowledge, the ALFF method has not previously been used to study the potential relationship between brain activity changes and clinical manifestations in patients with OF. This study aimed to explore the cerebral neural changes after orbital fracture using the fALFF technique ( Figure 5). The study found significantly lower fALFF values in the LACG and the RSTG in patients with OF (Figure 6). In previous studies, the fALFF method has been applied to a series of ophthalmological diseases, including normal-tension glaucoma (20), monocular blindness (21), retinal vein occlusion (22), diabetic retinopathy,  and nephropathy (23) ( Table 3), demonstrating its potential for clinical application. The anterior cingulate gyrus (ACG) is a functional area associated with many physiological functions, is located in the medial brain and passes longitudinally through the parietal lobe, and its main roles are in memory (24), action -outcome learning (25)(26)(27), emotion, and reward-related processing (28). The research of Hornak et al. (29) showed that, in some cases, the ACG plays an important part in voice and facial expression recognition, while Lane et al. (30)(31)(32) studied anterior cingulate injury in subjective emotional experience, and found that ventral ACG and Brodmann's area 9 may be activated during mood fluctuations. Based on the functions of the anterior cingulate gyrus, some researchers have explored its diagnostic value in Parkinson's disease (33), depression (34), and acute and chronic pain (35). In addition, a previous study has found that the prefrontal cingulate gyrus can respond to visual stimuli (36). In the present study, given the reduced visual responses in patients with OF, the results may indicate a compensatory mechanism for vision loss in patients with OF.
The superior temporal gyrus (STG) is a functional area of the brain located in the temporal lobe, closely related to emotional and speech processing (37,38). The STG is a component of the default mode network, which is inhibited during brain activity and excited during rest. Liu et al. (39) used the rsMRI-fALFF method to study the STG in depression. They found that the lower fALFF values of STG correlated greater reductions on the Hamilton rating scale for depression, and inferred that STG neural changes were closely related to the effect of early treatment for depression. In addition, Wang et al. (40) measured functional connectivity density in STG and found that abnormal connectivity is negatively correlated with the treatment effect. In the present experiment, the fALFF value of the right STG in patients with OF was significantly lower than that in healthy controls, and we speculate that this decrease may be a compensatory mechanism for the recovery of brain function in patients with OF. The results suggest that the fALFF value may be used as a reliable index to gauge therapeutic effects of clinical treatment. Moreover, it was discovered that, in the patients with OF, fALFF values in the RSTG were negatively correlated with anxiety and depression scores, which may indicate a selfregulation mechanism in this brain area, with brain function being temporarily inhibited ( Table 4).
This study has some limitations; one of which is the relatively small sample size, and the other is that the sample source was limited and not completely matched. Third, compared with the simple use of VEP, the use of pattern electro retino grams (PERGs) and pattern visual-evoked potentials (PVEPs) two checks can be more rigorous explanation of the problem. Therefore, future research should use larger and more closely matched samples to further clarify the neural changes in patients with orbital fractures, and to provide a more intuitive clinical efficacy index for treatment. In conclusion, this study has demonstrated that patients with OF have reduced fALFF values in specific cerebrum areas, indicating changes in spontaneous brain activity. Further research on the mechanism underpinning brain activity changes in patients with OF may be helpful to advance understanding of this condition.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Ethics Committee of Nanchang University. The patients/participants provided their written informed consent to participate in this study.