Altered resting-state cerebellar-cerebral functional connectivity in patients with end-stage renal disease

Abstract

Background

End-stage renal disease (ESRD) patients have functional and structural brain abnormalities. The cerebellum also showed varying degrees of damage. However, no studies on cerebellar-cerebral functional connectivity (FC) have been conducted in ESRD patients. This study aimed to investigate the changes in cerebellar-cerebral FC in ESRD patients and its relationship with neuropsychological and clinical indexes.

Methods

Resting-state functional magnetic resonance imaging and neuropsychological assessment were performed on 37 ESRD patients and 35 control subjects. Seed-based FC analysis was performed to investigate inter-group differences in cerebellar-cerebral FC. In addition, the relations of altered FC with the neuropsychological function and clinical indicators were analyzed in ERSD patients.

Results

ESRD patients exhibited alterations in cerebellar-cerebral FC involving the executive control network, default mode network, and affective-limbic network compared to control subjects (False discovery rate-corrected, p < 0.05). The altered cerebellar-cerebral FC was associated with the Montreal Cognitive Assessment Scale score (p < 0.05), and correlated with serum creatinine and uric acid levels within the ESRD group (p < 0.05).

Conclusions

The study indicates that cerebellar-cerebral FC is involved in the neural substrates of cognitive impairment in ESRD patients. The findings may provide clinically relevant new neuroimaging biomarkers for the neuropathological mechanisms underlying cognitive impairment of ESRD.

Keywords:

• End-stage renal disease
• cerebellum
• resting-state fMRI
• functional connectivity
• cognitive impairment

1. Introduction

Cognitive decline is prevalent in end-stage renal disease (ESRD) patients, with an incidence of 30–87%, primarily involving executive function, visuospatial function, attention, working memory, and processing speed [1,2]. Severe cognitive impairment can lead to a diminished quality of life, reduced adherence to treatment, and increased mortality [3,4]. Therefore, understanding the neuropathologic substrates of cognitive impairment in ESRD patients is essential for early intervention of cognitive impairment, guiding clinical treatment, and improving prognosis.

Brain imaging studies demonstrated that functional and structural brain abnormalities in ESRD patients have been detected in several brain regions, including the prefrontal lobe, inferior parietal lobe, cingulate gyrus, precuneus, temporal lobe, hippocampus/parahippocampal gyrus, and insula [5–9]. There were also abnormalities in the default mode network (DMN) [10–12], executive control network (ECN) [10,12], sensorimotor network (SMN) [12,13], salience network (SN) [10,14], and affective-limbic network (ALN) [15,16], which are composed of the brain regions mentioned above. The impairment of these brain regions or networks was related to abnormalities in overall cognitive function, working memory, visuospatial/executive function, attention, processing speed, sensorimotor, and emotional processing in ESRD patients. Although earlier studies have provided a neuroimaging basis for understanding the mechanisms of cognitive decline in ESRD patients [10,12], the neuropathologic substrate of ESRD remains unknown.

Several studies have proved that the cerebellum is implicated not only in motor control but also in emotional processing, memory, attention and executive function [17,18]. The cerebellum can communicate with the cerebral cortex via the cerebellar peduncle, dentate nucleus, pons, and thalamus. Moreover, it can be divided into different sub-regions connecting the cerebral ECN, SN, DMN, ALN, and SMN based on cerebellar-cerebral FC patterns [17–19]. These cerebellar-cerebral connections support that the cerebellum can contribute to motor control, cognitive and emotional processing. Recent neuroimaging studies also found the cerebellum or cerebellar-cerebral FC play a pivotal role in ensuring cognitive function in healthy individuals as well as contributing to clinical symptoms of neuropsychiatric and metabolic brain diseases, such as Alzheimer’s disease and mild cognitive impairment [20], hepatic encephalopathy [21], major depression disorder [22], and obsessive-compulsive disorder [23]. Previous brain imaging of ESRD patients also detected alteration in the white matter microstructure of the cerebellar peduncle [24–26], a reduction in the amplitude of low-frequency fluctuations in the cerebellar cortex [27], and changed functional and structural connectivity of the cerebellum [14,28,29]. The above studies indicated that ESRD patients have structural and functional abnormalities of the cerebellum. However, these cerebellar abnormalities were incidentally detected by analyzing the whole brain. Few studies have specifically paid attention to cerebellar abnormalities in ESRD, and there has been no study on cerebellar-cerebral FC in ESRD patients.

Resting-state fMRI is a method that indirectly infers information about brain activity by measuring Blood-Oxygen-Level Dependent (BOLD) signal. The temporal correlations of BOLD signals in different brain regions can be used to reflect brain FC [30]. The abnormalities in resting-state fMRI connectivity reflect alterations in the interactions among different brain regions [31]. Considering that DMN, ECN, ALN, and SMN are abnormal in ESRD, and that cerebellar sub-regions can identify the cerebellar-cerebral DMN, ECN, ALN, and SMN, we hypothesized that the cerebellum-cerebral FC is abnormal and related to cognitive impairment in ESRD patients. Therefore, the present study aimed to explore whether the cerebellar-cerebral FC is altered in ESRD patients with cerebellar sub-regions as seeds using resting-state fMRI, and further investigate the relationship between the altered FC, neuropsychological function, and clinical parameters in patients with ESRD.

2. Materials and methods

2.1. Subjects

Thirty-seven ESRD patients with peritoneal dialysis were recruited from The First Affiliated Hospital of Anhui Medical University, Hefei, China. The etiology of the ESRD was as follows: chronic glomerulonephritis (n = 24), IgA nephropathy (n = 4), polycystic kidney (n = 4), lupus nephritis (n = 2), anaphylactoid purpura (n = 2), nephrotic syndrome (n = 1). The recruitment criteria for ESRD patients were as follows: dialysis duration was not less than three months in patients; aged 20–65 years; patients with stable disease without dyspnea, heart failure, nausea, fatigue, or other symptoms. Exclusion criteria were the presence of craniocerebral trauma, tumor, stroke, or other neurological diseases, history of alcohol or drug abuse, previous history of neuropsychiatric illness, central nervous medication is taken during the past 3 months, head movement translation or rotation > 2.0 mm or 2.0° during MRI scan, or the presence of contraindications to MRI examination. We recruited 35 age- and gender-matched control subjects using the same exclusion criteria as that for the patients.

2.2. Neuropsychological tests

Participants underwent a series of neuropsychological tests within 24 h before the MRI scan. The Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment Scale (MoCA), and Digit Symbol Substitution Test (DSST) were performed to assess overall cognitive function, working memory, visuospatial/executive function, attention, and processing speed. The Hamilton Depression Rating Scale (HAMD) and Hamilton Anxiety Rating Scale (HAMA) were performed to evaluate depressive and anxiety symptoms.

2.3. Laboratory examinations

All patients received laboratory blood biochemical examinations within 24 h before the MR examination, including GFR, creatinine, urea nitrogen, uric acid, hemoglobin, sodium, potassium, calcium and phosphorus. No blood biochemical examination was performed in the control group.

2.4. MRI data acquisition

Please refer to the Supplementary Material for details on magnetic resonance data acquisition.

2.5. Image preprocessing

DICOM images were converted to NIFTI format using dcm2nii software. The functional images at the first five volumes were discarded to achieve signal equilibrium and subjects’ adaptation to the imaging conditions. Afterwards, based on the MATLAB software (MATLAB R2018a, Math Works, USA), the default Montreal Neurological Institute (MNI) pipeline in the CONN v19 toolbox (https://www.nitrc.org/projects/conn) was used to further preprocess the MRI data, including slice timing correction, head motion assessment and correction, segmentation and normalization of structural and functional images in the MNI stereotactic space, and spatial smoothing with an isotropic Gaussian kernel with 6-mm full width at half maximum. Subsequently, the covariates (including six head-motion parameters, cerebrospinal fluid, white matter, and other high-frequency physiological noises) and linear drift were removed. Finally, the datasets were temporally filtered in the 0.01–0.08 Hz band.

2.6. Definition of the cerebellar seeds

Thirteen cerebellar seeds (Supplemental Table 1) were selected to construct the cerebellar-cerebral FC. These seeds are able to identify cerebellar-cerebral DMN, ECN, ALN and SMN [17,19], and have been used in previous studies [22,23]. Each seed was created as a 6-mm radius sphere with these coordinate points (Supplemental Table 1) as the center in standard MNI space.

2.7. Statistical analysis

Demographic and neuropsychological data were analyzed using SPSS statistics version 22.0. A two-sample independent t-test or Chi-squared test was performed to obtain the differences in age, years of education, neuropsychological test scores, and gender between the two groups; p < 0.05 was considered statistically significant.

The cerebellar-cerebral FC was calculated through first-level seed-to-voxel FC analyses using the CONN toolbox. The Pearson correlation coefficients (r) were computed to establish the relationship between the average BOLD time series of each seed and those of each voxel located in the cerebral cortex. The r value serves as an indicator of the degree of FC. Fisher’s r-to-z transformation was used to improve the normal distribution of the data. Then, the cerebellar-cerebral FC maps of each subject were obtained for statistical analysis. A second-level analysis model was employed to ascertain the differences in cerebellar-cerebral FC between the two groups. This was achieved through the use of multivariate ANCOVA covariates control, with age, gender, and education attainment serving as covariates (voxel-level of p < 0.005, False discovery rate (FDR)-corrected, cluster-level of p < 0.05, FDR-corrected, and cluster size > 40 voxels). The cerebellar-cerebral functional connectivity (FC) image, exhibiting notable inter-group distinctions, was projected onto the brain template furnished by the CONN toolbox. Subsequently, the image was partitioned into distinct anatomical regions within the cortex.

The altered cerebellar-cerebral FC values (Z-scores) were then extracted, and Pearson correlation analysis was employed to determine the correlations between the Z-scores and neuropsychological and clinical indexes (p < 0.05) in the ESRD group. Due to the exploratory nature of these correlation analyses, the correction for multiple testing was not performed in this study.

3. Results

3.1. Demographic, clinical, and neuropsychological features

Table 1 displays the demographic, neuropsychological assessment, and clinical indicators data of the ESRD and control groups. No statistically significant differences in age, gender, and educational attainment between the two groups. The neuropsychological performance levels in the ESRD group were significantly different compared to the control group. Additionally, MoCA visuospatial/executive function, attention and delayed recall scores differed significantly between the two groups (Supplemental Table 2).

Table 1. Demographic, clinical and neuropsychological data for the ESRD and control groups.

Variables	ESRD patients (n = 37)	Control subjects (n = 35)	p-value
Age (year)	46.62 ± 9.15	48.34 ± 9.35	0.433
Gender (male/female)	13/24	17/18	0.248
Education (years)	6.91 ± 2.98	8.31 ± 3.14	0.057
Disease duration (month)	77.08 ± 80.46	−	−
Dialysis duration (month)	25.14 ± 28.32	−	−
Creatinine (μmol/L)	956.38 ± 281.06	−	−
Urea nitrogen (mmol/L)	18.83 ± 5.37	−	−
Uric acid (μmol/L)	428.78 ± 78.20	−	−
Hemoglobin (g/L)	94.00 ± 17.05	−	−
Sodium (mmol/L)	138.35 ± 4.51	−	−
Potassium (mmol/L)	3.95 ± 0.58	−	−
Calcium (mmol/l)	2.23 ± 0.20	−	−
Phosphorus (mmol/l)	1.63 ± 0.58	−	−
MMSE	26.41 ± 3.21	28.69 ± 1.64	<0.001
MoCA	23.30 ± 3.90	25.43 ± 3.05	0.012
DSST	35.00 ± 15.4	42.95 ± 13.58	0.024
HAMD	3.86 ± 3.58	1.46 ± 1.48	<0.001
HAMA	4.76 ± 3.30	2.25 ± 1.48	<0.001

ESRD: end-stage renal disease; MMSE: Mini-Mental State Examination; MoCA: Montreal Cognitive Assessment Scale; DSST: Digit Symbol Substitution Test; HAMD: Hamilton Depression Rating Scale; HAMA: Hamilton Anxiety Rating Scale.

3.2. Inter-group differences in cerebellar-cerebral FC

The cerebellar-cerebral FC differed significantly between ESRD and control groups (Table 2, Figure 1). The right cerebellar Crus I_Exec1 showed decreased FC with the left middle frontal gyrus and increased FC with the right insula cortex (Figure 1(A)). The right cerebellum Crus II_Exec2 showed decreased FC with the bilateral middle frontal gyrus, the left angular gyrus, the left superior frontal gyrus, and increased FC with the right insular cortex and the right hippocampus (Figure 1(B)). The left cerebellar Crus II_Exec2 showed decreased FC with the right angular gyrus, the right middle frontal gyrus, and the left superior frontal gyrus (Figure 1(C)). The left Lobule VI_Exec3 showed decreased FC with the bilateral middle frontal gyrus (Figure 1(D)). The right cerebellar Crus I_DMN showed decreased FC with the left superior frontal gyrus, the left medial frontal gyrus, the left angular gyrus, increased FC with the right insula cortex, the right precuneus/posterior cingulate gyrus (Figure 1(E)); the right cerebellar Lobule VI_Aff showed decreased FC with the bilateral middle frontal gyrus and the left supramarginal gyrus (Figure 1(F)).

Figure 1. Differences in cerebellar-cerebral FC between the ESRD and the control groups (p < 0.05, FDR-corrected). altered FC of the R Crus I_Exec1 in ESRD (a). altered FC of the R Crus I_Exec2 in ESRD (B). altered FC of the L Crus II_Exec2 in ESRD (C). altered FC of the L Crus I_Exec3 in ESRD (D). altered FC of the R Crus I_DMN in ESRD (E). altered FC of the R Lobule VI_Aff in ESRD (F). L, left; R, right; FC, functional connectivity; ESRD, end-stage renal disease; FDR, false discovery rate.

Table 2. Alterations in the cerebellar–cerebral FC in the ESRD group compared with the control group.

Seeds	Brain regions	Voxels	Peak’s MNI coordinate
			X	Y	Z	t-value
R Crus IExec1	L middle frontal gyrus	356	−46	12	50	−7.44
		105	−28	54	−4	−6.32
	R insular cortex	85	34	−16	−2	6.56
R Crus IIExec2	L middle frontal gyrus	985	−36	40	−14	−6.52
	L superior frontal gyrus	1280	−22	26	58	−7.52
	L angular gyrus	226	−42	−54	40	−5.60
	R insular cortex	163	34	−14	−2	7.64
	R middle frontal gyrus	158	40	50	−8	−5.85
	R hippocampus	332	24	−38	−8	5.30
L Crus IIExec2	R angular gyrus	240	40	−66	50	−6.23
	R middle frontal gyrus	195	28	24	48	−6.21
	L superior frontal gyrus	89	−24	16	64	−5.85
L Lobule VIExec3	L middle frontal gyrus	105	−28	50	28	−6.76
	R middle frontal gyrus	65	30	44	28	−5.63
R Crus IDMN	L superior frontal gyrus	1480	−2	44	50	−7.58
	L medial frontal gyrus	405	2	50	−18	−6.31
	R precuneous/posterior cingulate gyrus	212	12	−30	44	7.53
	L angular gyrus	133	−46	−58	32	−5.72
	R insular cortex	80	34	−16	−2	6.14
R Lobule VIAff	L middle frontal gyrus	256	−32	56	4	−6.50
		135	−36	10	40	−6.00
	R middle frontal gyrus	102	34	46	−2	−6.00
	L supramarginal gyrus	72	−48	−48	48	−5.65

ESRD, end-stage renal disease; Exec, executive; DMN, default mode network; Aff, affective; Mot, motor. R, right, L, left. MNI, Montreal Neurological Institute.

3.3. Correlation between altered cerebellar-cerebral FC and cognitive function

As illustrated in Figure 2. MoCA scores showed negative correlations with FC between the right cerebellum Crus I_Exec1 and the left middle frontal gyrus (Figure 2(A)), the right cerebellum Crus II_Exec2 and the left middle frontal gyrus (Figure 2(B)), the right cerebellum Crus II_Exec2 and the left angular gyrus (Figure 2(C)), the left cerebellum Crus II_Exec2 and the right angular gyrus (Figure 2(D)), the left cerebellum Crus II_Exec2 and the right middle frontal gyrus (Figure 2(E), the left cerebellar Lobule VI _Exec3 and the left middle frontal gyrus (Figure 2(F)), the left cerebellar Lobule VI _Exec3 and the right middle frontal gyrus (Figure 2(G)). Significant correlations existed between altered cerebellar-cerebral FC and MoCA visuospatial/executive function scores and MoCA attention scores (Supplemental Figure 1). There were no statistically significant correlations observed between altered cerebellar-cerebral FC and other neuropsychological scores.

Figure 2. Correlations between altered cerebellar-cerebral FC and MoCA scores in the ESRD group. Significantly negative correlations between MoCA scores with the FC between the right cerebellum Crus I_Exec1 and the left middle frontal gyrus (a), the right cerebellum Crus II_Exec2 and the left middle frontal gyrus (B), the right cerebellum Crus II_Exec2 and the left angular gyrus (C), the left cerebellum Crus II_Exec2 with the right angular gyrus (D), the left cerebellum Crus II_Exec2 with the right middle frontal gyrus (E), the left cerebellar Lobule VI _Exec3 and the left middle frontal gyrus (F), the left cerebellar Lobule VI_Exec3 and the right middle frontal gyrus (G). L, left; R, right; FC, functional connectivity; ESRD, end-stage renal disease; MoCA, Montreal cognitive assessment scale.

3.4. Correlation between altered cerebellar-cerebral FC and clinical variables

As depicted in Figure 3. Serum creatinine level was negatively correlated with the FC between the left cerebellum Crus II_Exec2 and the right middle frontal gyrus (Figure 3(A)), as well as the left cerebellum Crus II_Exec2 and the left superior frontal gyrus (Figure 3(B)), Uric acid level showed negative correlation with the FC between the right Crus I_DMN and the right insular cortex (Figure 3(C)). No significant correlations were found between altered cerebellar-cerebral FC and other clinical variables.

Figure 3. Correlations between altered cerebellar-cerebral FC and clinical variables in the ESRD group. Serum creatinine levels showed negative correlation with the FC between the left cerebellum Crus II_Exec2 and the right middle frontal gyrus (a), and the left cerebellum Crus II_Exec2 and the left superior frontal gyrus (B), uric acid levels showed negative correlation with the FC between the right Crus I_DMN and the insular cortex (C). L, left; R, right; FC, functional connectivity; ESRD, end-stage renal disease.

4. Discussion

To our knowledge, this is the first report to explore cerebellar-cerebral FC in ESRD patients. The study found that the FC in cerebellar-cerebral ECN, DMN, and ALN were changed in ESRD patients. Part of abnormal FC was correlated with the serum creatinine and uric acid levels. In addition, the FC between cerebellar sub-regions and the middle frontal gyrus and the angular cortex were related to the MoCA scores within the ESRD group. The results indicate that the cerebellar-cerebral FC was disturbed by impaired renal function, and may be involved in the neuropathologic basis of cognitive impairment in the ESRD population.

In the study, ESRD patients displayed altered FC between the cerebellum and the prefrontal lobe (superior, middle, and medial frontal lobe), inferior parietal lobe (the angular and supramarginal gyrus), insula, hippocampus, and precuneus/posterior cingulate gyrus. These alterations were mainly located within DMN, ECN, and ALN. Previous studies have identified impairments in these brain networks [10,12,14], which align with the findings of our study. The frontoparietal regions are important components of the DMN and ECN, and the cerebellar-frontoparietal network is involved in various higher cognitive abilities [32,33]. The prefrontal cortex is associated with executive functions [33,34]. Executive dysfunction is one of the most common cognitive dysfunctions in ESRD [35]. There was a similar finding in our study. Executive dysfunction may be associated with extensively decreased FC between the cerebellar sub-regions and the prefrontal cortex. Additionally, some scholars considered that executive dysfunction in patients with chronic kidney disease may be attributed to a decreased thickness of the frontal cortex [1]. The inferior parietal lobule is one hub of the DMN. Furthermore, previous studies showed that ESRD patients displayed changes in low-frequency amplitude and degree centrality in the inferior parietal lobule [36,37]. The reduced FC between the cerebellum and the inferior parietal lobule may impair the visuospatial, executive, and memory function systems of ESRD patients. The FC in cerebellar Crus I_Exec1, Crus II_Exec2, and Lobule VI_Exec3 with the middle frontal gyrus, and the angular gyrus were correlated with MoCA total scores, visuospatial/executive function scores and attention scores. Furthermore, the cerebellar Crus I, Crus II, and Lobule VI are involved in attention, working memory and visuospatial/executive function [38]. The results indicated that there was a direct relationship between cerebellar-cerebral FC and cognitive decline in ESRD patients. The prefrontal and inferior parietal lobules are also implicated in emotional regulation and processing [39,40]. The reduced FC in the right cerebellar Lobule VI_Aff with the middle frontal and left supramarginal gyrus was detected in this study, which might be associated with mood disorders like depression or anxiety in ESRD patients. The increased FC in the Crus _Exec and Crus _DMN with the hippocampus, insula, and precuneus/posterior cingulate gyrus were detected in ESRD patients. The alteration in the FC between the cerebellum and hippocampus may influence memory, spatial navigation learning, and emotional behaviour [41]. The insula is involved in cognitive control, attention regulation, and planning decision-making, also an essential player in switching between the DMN and ECN during cognitive processing [14]. The precuneus/posterior cingulate gyrus are the main nodes of the DMN, and participate in episodic memory and visuospatial processes [42]. However, the exact mechanism of increased FC is unclear, and there may be two following explanations. First, the increased FC is a fundamental response to neurological disruption and may be a compensatory or adaptive response to disease-related pathology [43]. Second, evidence suggests that brain dysfunction in ESRD might be associated with an imbalance of excitatory and inhibitory effects [44]. Uremic neurotoxins can cause a disproportion of γ-amino butyric acid (GABA) and N-methyl-D-aspartate, thus shifting cortical circuitry toward greater excitability or failure of inhibitory influences that may be responsible for increased FC [44,45].

There were overlapping brain regions in the altered cerebellar-cerebral ECN, DMN, and ALN, including the medial frontal gyrus, middle frontal gyrus, angular gyrus, hippocampus and insula. The finding supported information exchange among these networks and provided new insight into the abnormal interactions between the affective and cognitive control networks in ESRD patients [15,16]. Additionally, The result also indicated that the cerebellum is a crucial node in the brain network and plays a pivotal role in transmitting information within the network in ESRD patients.

The present study observed that serum creatinine and uric acid levels were correlated with the cerebellum-cerebral FC in ESRD patients, suggesting that cerebellum-cerebral FC was disturbed by renal impairment. We observed that the higher level of creatinine, the poorer cerebellar–cerebral FC in ESRD patients, which is consistent with the previous study [11,36]. Creatinine is an indicator of renal function, not a neurotoxin, thus it might be a surrogate marker of putative neurotoxin accumulating in ESRD patients [11]. The results indicated the progression of renal impairment may aggravate brain function abnormality. Uric acid is a uremic toxin with antioxidants (mainly in plasma) and oxidation (mainly in cells), which can play neuroprotective and neurotoxic effects [1,46]. In our study, the uric acid level showed a negative correlation with the FC between the right Crus I_DMN and the insular cortex, which indicated that the neurotoxic effect of uric acid was probably dominant in ESRD patients. The findings provided a neuroimaging basis to understand the relationship between the clinical relevance risk factors and brain damage in ESRD.

The study has several limitations. First, the sample size of patients is relatively small, the impact on peritoneal dialysis will be explored with large sample sizes. Second, hypertension, hyperlipidemia, diabetes, and electrolyte disturbances are common complications of ESRD. It is necessary to consider the influence of these factors on brain abnormalities in patients with ESRD in further studies. Third, the correction for multiple testing was not used in the correlation analyses due to the small sample size. Finally, this study is a cross-sectional study. A cohort study may provide further evidence to evaluate clinical treatment effects in future research.

Conclusion

This study suggests that the cerebellar-cerebral FC is involved in the neural substrates of cognitive impairment in ESRD population. The findings provide clinically relevant neuroimaging biomarkers to understand further the neuropathological mechanisms underlying cognitive impairment of ESRD.

Author contributions

JF contributed to experimental design, data acquisition, data handling, and data interpretation, and wrote the article. FZ, JLZ, YRL and YYM were involved in clinical evaluation and data acquisition. HBW and XMQ contributed to the study design, data interpretation and revised the manuscript.

Ethical approval

This study involving human participants was reviewed and approved by the biomedical ethics committee of Anhui Medical University.

Patient consent

All participants provided their written informed consent to participate in this study.

Consent to publish

The manuscript has been read and approved for submission by all (co-)authors.

Acknowledgments

We also would like to thank Freescience for English language editing.

Disclosure statement

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.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This research was supported by Natural Science Foundation of Anhui Province (No. 1908085MH245) and Natural Science Research Project of Anhui Universities (No. KJ2018A0493)