Functional connectivity density alterations in middle‐age retinal detachment patients

Abstract Objective Middle‐age patients with retinal detachment (RD) exhibit a loss of visual information, and previous studies of functional magnetic resonance imaging (fMRI) have demonstrated abnormal spontaneous activity in the RD brain. Therefore, this study assessed changes in local functional connectivity density (lFCD) and long‐range functional connectivity density (longFCD) in middle‐age RD patients during resting‐state FC. Methods In total, 32 middle‐age patients with RD (18 men and 14 women), and 32 age‐, sex‐, and education‐matched normal controls (NCs) (18 men and 14 women) were recruited and underwent functional magnetic resonance examination in the resting state. Two‐sample t test was performed to compare lFCD and longFCD between groups. Receiver operating characteristic (ROC) curves were generated to distinguish middle‐age RD patients from NCs. Results Compared with NCs, middle‐age RD patients demonstrated increased lFCD values in the right inferior temporal gyrus, and increased longFCD values in the bilateral inferior frontal gyri, left superior and middle frontal gyrus, bilateral inferior temporal gyri, and left cerebellum posterior lobe. Middle‐age RD patient exhibited decreased lFCD values in the left cuneus, right lingual gyrus, and left middle frontal gyrus. They also had lower longFCD values in the left lingual gyrus and left inferior occipital gyrus. ROC curve analysis showed excellent accuracy of the specific areas under the curve. Conclusion Our results reveal that middle‐age RD patients exhibited variations of binarized lFCD and longFCD in specific brain areas, which provides insight into the pathological mechanism of RD patients with acute visual loss.


| INTRODUC TI ON
Retinal detachment (RD) (Figure 1) refers to separation between the neural retina and retinal pigment epithelium; this leads to vision loss and increases the physical distance between photoreceptor cells and the retinal pigment epithelium and choriocapillaris (Fisher, et al., 2010). The main risk factors for middle-age RD are uveitis (Amer, Nalci, & Yalcindag, 2017), high myopia (Baba et al., 2003), vitreous detachment (Fincham et al., 2018), and ocular trauma (Hoogewoud et al., 2016). Foveal detachment can rapidly damage central visual acuity (VA), and surgery is the main treatment (García-Arumí et al., 2013). RD is typically diagnosed with B-ultrasound and optical coherence tomography (OCT). Compared with B-ultrasound, OCT is superior for observing the microstructural retina changes. The regional homogeneity method has been used to observe changes in brain neural homogeneity in middle-age RD patients to understand their relationships with clinical features (Huang et al., 2017).
Resting-state functional connectivity (rs-FC) analysis is a useful method to evaluate spontaneous functional organization though FC, and it gauges the temporal correlation between the time series of the blood oxygen level-dependent (BOLD) signals of two brain regions (Tomasi et al., 2010). rs-FC has been successfully applied to study depression (Klumpp, Hosseini, & Phan, 2018;Maglanoc et al., 2019), Alzheimer's disease (Hata et al., 2016), Parkinson's disease (Hu et al., 2017), and primary open-angle glaucoma (Dai et al., 2013).
Recently, we have observed close correlation between the default mode network functional connections and visual function impairments of RD patients (Su et al., 2018). However, alterations in the local FC density (lFCD) and long-range FC density (longFCD) have not been revealed. To investigate the effect of middle-age RD patients on visual pathway and cortex, we applied rs-FC to assess differences in local FC density (lFCD) and long-range FC density (longFCD). We hypothesized that the lFCD and longFCD would be altered in middle-age RD patients and that the spatial distribution of abnormal lFCD or longFCD and the abnormal pattern (increased or decreased) of lFCD or longFCD would be different in middle-age RD patients compared with normal controls. Identifying unique changes in middle-age RD patients could clarify the pathological mechanism and guide new treatment strategies.

| Subjects
We recruited 32 middle-age RD patients (18 men and 14 women) who were treated in the First Affiliated Hospital of Nanchang University and examined by a single experienced ophthalmologist. The inclusion criteria were as follows: (1) middle-age RD occurred within 1 month, (2) macular detachment, and (3) no comorbid ocular diseases (e.g., strabismus or amblyopia).The exclusion criteria were as follows: (1) systemic disease (e.g., cardiovascular disease or cerebral disease) and (2) psychiatric disorders (e.g., delusional or depressive disorder).
All research methods were approved by the committee of the medical ethics of the First Affiliated Hospital of Nanchang University, and participants were explained the intent, method, and potential risks, and they signed an informed consent form.

| MRI data acquisition and data analysis
MRI was performed on a 3.0 Tesla (T) system (Magnetom TRIO, Siemens Medical; Erlangen, Germany). High-resolution T1-and T2weighted imaging sequences were implemented with the follow- To balance the MRI scanning signal, the first 10 image was deleted. Based on MATLAB 2010a (MathWorks, Natick, MA, USA), the rest of the data preprocessing was performed by Data Processing & Analysis for Brain Imaging (DPABI 2.1, http://rfmri.org/DPABI) toolbox, including slice timing, head-motion correction, and spatial normalization. Participants who had more than 1.5 mm maximum translation in x, y, or z directions and 1.5° degree of motion rotation were excluded from further analyses. The Friston 24 head-motion parameter model was used to regress out head-motion effects, based on recent work showing that higher-order models benefit from their removal (Satterthwaite et al., 2013;Yan et al., 2013).
Linear regression was performed to remove other sources of spurious covariates and their temporal derivatives, including the global mean signal and the white matter and cerebrospinal fluid signals.
Functional MRI images were normalized spatially to the Montreal Neurological Institute (MNI) space and resampled at a resolution of 3 × 3×3 mm 3 after head-motion correction. Then, the time series for each voxel was temporally bandpass-filtered (0.01-0.1 Hz) and linearly detrended to reduce low-frequency drift, and physiological high-frequency breathing and heart noise.

| Calculation of lFCD and longFCD Maps
FCD maps contained long-and short-range data for each individual, the short-range FCD equates with local FCD, and all of them were calculated in a gray matter mask. Based on the Pearson correlations between the time course of a given voxel and that of other voxels, the number of functional connections of a given voxel was considered as a degree of a node in a binary graph. The detailed procedure for computing lFCD and longFCD is published elsewhere (Faivre et al., 2016). First, we defined FC in the whole brain between a given voxel with every other voxel, with a correlation threshold of r > 0.25 (Hata et al., 2016). We employed the threshold of r = 0.3 to calculate the FCD maps. Second, the lFCD and longFCD were defined on the basis of the neighborhood strategy, which means that voxels that met the correlation threshold of r = 0.30 inside and outside their neighborhood (radius sphere ≤6 mm and >6 mm) were defined as lFCD and longFCD, respectively. To improve normality, we divided the lFCD and longFCD maps by the mean value for the whole brain of each subject and converted the values to Z scores. Finally, SPM8 software was used to generate the lFCD and longFCD maps, which were spatially smoothed with a Gaussian kernel of 6 × 6×6 mm 3 full width at half maximum.

| Ophthalmic testing
The international standard logarithmic VA chart was utilized for visual test analysis. Middle-age RD patients were diagnosed by Bultrasound and spectral domain OCT.

| Statistical analysis
Independent-sample t tests were performed using SPSS version 16.0 (SPSS Inc, Chicago, IL, USA) to analyze the cumulative clinical measurements, including RD duration. Differences were considered significant at p < .05.
One-sample t test was performed to extract the lFCD and longFCD value results across the subjects within each group and to examine the significant intragroup distribution. Differences were considered significant at p < .05. Statistical analysis of general linear model was carried out using the SPM8 toolkit. The twosample t tests were used to examine the differences of the FCD maps between the middle-age RD groups and the NCs (two-tailed, voxel-level p < .01, Gaussian random field theory (GRF) correction, cluster-level p < .05).
Receiver operating characteristic (ROC) curves were generated to distinguish the average lFCD from longFCD values in the various brain regions between the two groups. Differences were considered significant at p < .05. Abbreviations:N/A, not applicable; NCs, normal controls; RD, retinal detachment; VA, visual acuity.
*p < .05, independent t tests comparing two groups, Data shown as mean ± standard deviation.

| Demographics and visual measurements
There were no apparent differences in weight (p = .87) or age (p = .89), but we observed significant differences in best-corrected VA-right (p = .001) and best-corrected VA-left (p = .003) between the two groups. The mean standard deviation of RD duration was 23.12 ± 11.20 days. Details are presented in Table 1.

| FCD analysis
Compared with the NC group, middle-age RD patients demonstrated

| ROC curve analysis
We hypothesized that the different lFCD and longFCD values could be useful diagnostic markers to differentiate between middle-age RD group and NCs. ROC curve analysis was performed to test this assumption. The mean FCD values of distinct brain areas were collected and analyzed. The accuracy is low if the area under the curve (AUC) is 0.5-0.7, and when the AUC is 0.7-0.9, it means the accuracy is excellent. The individual AUCs of the binarized lFCD values for altered brain regions were as follows: right ITG (0.950, p < .001), left Notes: Between-group differences in binarized lFCD thresholded at r = 0.3. We used thresholds of two-tailed voxel-wise p < .01 and cluster-level p < .05, corrected for multiple comparisons by AlphaSim to determine the significant group differences.

| D ISCUSS I ON
Our results showed alterations in lFCD and longFCD in specific brain regions of patients with middle-age RD (Figures 5 and 6). These changes may result in decreased VA, visual field defects, and poor resolution of external object size or color. The human visual system has highly efficient visual information processing. The pathway begins with light stimulation, which induces the nerve impulse to the visual formation.
The visual information impulse is further transferred to the primary (V1) and secondary (V2) visual cortices, which consist of 19 and 18 zones, respectively. Previous studies of fMRI have demonstrated abnormal spontaneous activity in the RD patients. Decreased regional homogeneity values and abnormal rs-FC in DMN subregions were observed in RD patients (Huang et al., 2017;Su et al., 2018), these reflected synchronous dysfunction of brain activity and might hindered the information connection of specific brain regions.
Compared with NCs, we found significant increases in FCD in some V2 areas (Brodmann area 18) including the left cuneus and right lingual gyrus for lFCD, and the left IOG and left lingual gyrus for longFCD. It is well known that the bilateral calcarine cortex in the occipital lobe receives input from the optic radiation, and the calcarine sulcus divides the interior of the posterior cerebral hemisphere into the cuneus and lingual gyrus. There is a strong theoretical and anatomical basis for decreased FCD in V2 due to RD. This result demonstrates that loss of visual information affects brain activity, which could help guide the development of RD treatments. It also helps explain the effect of losing stereoscopic vision in the corresponding functional areas of the brain. In addition to focusing on vision recovery, we should also observe and follow up brain function changes.  (Bartsch et al., 2017) reported that attention to color is indexed by a sequence of modulations of the MRI brain response that are generated in the ventral extrastriate visual cortex. We also observe decreased FCD in these brain regions. To some extent, we believe these effects contribute to early disease symptoms.

F I G U R E 3
Similarly, our brain processes numerosity, which is one of the basic visual attributes needed to acquire a detailed portrayal of the outside world (Dakin, Tibber, Greenwood, Kingdom, & Morgan, 2011;Durgin et al., 2008). Roggeman, Santens, Fias, and Verguts (2011) captured fMRI responses in the IOG and provided cogent proof for an occipitoparietal pathway conveying numerosity information. Another recent report (Fornaciai, Brannon, Woldorff, & Park, 2017) further demonstrated that numerosity processing begins at least by V2/V3. Compared with NCs, we observed lower longFCD in left IOG and other areas of V2 in RD patients, which may affect the effectiveness and accuracy of numerosity processing.
Determining whether the effects of uni-or bilateral RD are the same will require further studies.
We also observed the decreased longFCD in the left CPL. This brain region has a vital function in elaborate motor coordination; for example, it can restrain involuntary movement though inhibitory neurotransmitters, especially GABA (Campbell, Meek, Zhang, & Bell, 2007). The CPL mainly receives input from the brainstem (reticular formation and inferior olivary nucleus) and cerebral cortex (Ramnani et al., 2012). Information received from sensorimotor associative cortices passes several parts of the brainstem and enters the retral section of the cerebellum through the medial peduncle (Kelly et al., 2003). Furthermore, there are anatomical relationship between the CPL and brain regions involved in higher cognitive functions. Middle-age RD patients have difficulty performing fine movement because of decreased VA and visual field defect in the affected eye. The reduced longFCD values in the CPL of middle-age RD patients may indicate reduced fine motor function.
Middle-age patients with RD exhibited enhanced lFCD in the right ITG and higher longFCD in the bilateral ITG. The ITG is located below the MTG and is connected to the IOG, which is involved in ventral stream processing. This area is related to the portrayal of intricated object characteristics and the face perception (Haxby, Hoffman, & Gobbini, 2000). As the last location of the ventral cortical visual system, the ITG may be more active due to compensation for reduced VA in middle-age RD patients.
The SFG makes up about a third of the frontal lobe. Based on fMRI experiments, Goldberg et al. proposed that the SFG was involved in self-awareness and coordinated sensory system movements (Goldberg, Harel, & Malach, 2006). The SFG contributes to higher cognitive functions, especially working memory. Brodmann area 11 is involved in decision making (Rogers et al., 1999), processing rewards (Kringelbach et al., 2004), planning, encoding new information into long-term memory (Frey et al., 2000), and reasoning. Recovery of VA after RD takes a long time, even with active treatment, and patients often experienced persistent loss of VA, Notes: Between-group differences in binarized longFCD thresholded at r = 0.3. We used thresholds of two-tailed voxel-wise p < .01 and cluster-level p < .05, corrected for multiple comparisons by AlphaSim to determine the significant group differences. The MFG plays an important role in cognition and attention function (Chang et al., 2011). Decreased regional homogeneity values in the MFG were observed in RD patients, and this reflected local synchronization dysfunction in specific brain region (Huang et al., 2017). We found decreased lFCD in the left MFG. Compared with NCs, middle-age RD patients lost different degrees of visual information. This could negatively impact cognition or attention, but the specific mechanism needs further study. We also observed enhanced longFCD in the bilateral IFG. Studies (Cha et al., 2016;Vanderhasselt, Kuhn, & De Raedt, 2013) suggest that the IFG may have a more complex role in emotion regulation; specifically, it is involved in evaluating stimulus meaning, which then informs the ventromedial prefrontal cortex to inhibit the amygdala. Due to the negative effects of visual information loss, RD patients may experience hypoemotivity or depression that increases FCD. The IFG is also considered a language region (Hobson et al., 2018). The left IFG has been identified as a neural crossroads between different types of information that are equally necessary for differentiating abstract and concepts. The right IFG plays a role in attentional switching (Hampshire et al., 2006) and appears to interact with areas of the temporal lobe implicated in memory (Aron et al., 2014). One group (Hampshire, Chamberlain, Monti, Duncan, & Owen, 2010) also reported that the right IFG is involved in inhibitory control and attentional control. We speculate that the adverse effects experienced by middle-age patients with RD may influence their ability to deal with language and stimulate the brain to divert attention to overcome  Abbreviations: lFCD, local functional connectivity density; longFCD, long-range functional connectivity density; ROC, receiver operating characteristic these effects, but studying this requires larger samples, more clinical data, and longer follow-up.

| LI M ITATI O N S
There are some limitations in this study. First of all, the emotion scales and cognitive assessments were not collected, previous studies have demonstrated abnormal spontaneous activity in patient with depression and cognitive impairment, and therefore, patients' psychological and cognitive assessment data should be concerned and collected. Secondly, longitudinal studies are needed to investigate the influence of course of RD patients on lFCD and longFCD. In addition, we can assess the alterations of lFCD and longFCD in cured RD patients and compare them with that before any treatments, so as to observe whether the effect of FCD in RD patients is reversible. Finally, due to the small sample size of the study, the patients with RD were not further divided into groups according to right/left eye and area of RD, and we will include more behavioral performance data (e.g., electroretinogram and visual evoked potential) for analysis, larger samples should be included in future studies.

F I G U R E 5
The binarized lFCD results of brain activity in the RD group. Notes: Compared with the NCs, the binarized lFCD of the following regions in the RD group was as follows: 1-right inferior temporal gyrus (t = 4.4487), 2-left cuneus (t = −4.0319), 3-right lingual gyrus (t = −3.8358), and 4-left middle frontal gyrus (t = −3.4367). Abbreviations: lFCD, local functional connectivity density; RD, retinal detachment; NCs, normal controls F I G U R E 6 The binarized longFCD results of brain activity in the RD group. Notes: Compared with the NCs, the binarized longFCD of the following regions in the RD group was as follows: 1-left superior frontal gyrus (

| CON CLUS ION
Middle-age RD patients exhibited variations in binarized lFCD and longFCD in brain regions. Our results provide insight into the pathological mechanism of acute visual information loss in patients with RD.

CO N FLI C T O F I NTER E S T S TATEM ENT
This was not an industry-supported study. The authors report no conflicts of interest in this work.

E TH I C A L A PPROVA L A N D I N FO R M ED CO N S ENT
All procedures performed in studies involving human participants were in accordance with the ethical standards of the First Affiliated Hospital of Nanchang University ethical committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

ACK N OWLED G M ENTS
All authors participated equally in this study. Thanks to all the staffs in the department of Jiangxi Province Ocular Disease Clinical Research Center, First Affiliated Hospital of Nanchang University, for their technical support.

AUTH O R S' CO NTR I B UTI O N S
Shao Y and Lin Y designed and drafted the manuscript. Zhu P, Su T, and Zhou X analyzed the data. Li B, Shi W, Lin Q, Min Y, Yuan Q, and Ye L collected the data. Zhou Q revised the manuscript.

PEER R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1002/brb3.1783.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.