Altered Small-World Topological Properties of Functional Brain

Network in Patients with OSA

Ailin Hou

1,2,*

, Xueming Pang

and Quan Zhang

3,*

School of Precision Instrument & Opto-Electronics Engineering, Tianjin University, Tianjin 300072, China

School of Medical Imaging, Tianjin Medical University, Tianjin 300203, China

Department of Radiology, Characteristic Medical Center of Chinese People's Armed Police Force, Tianjin 300162, China

Keywords:

Obstructive Sleep Apnea, Resting-State, Functional Magnetic Resonance Imaging, Graph Theory, Small-

World Property.

Abstract: Obstructive sleep apnea (OSA) is a common disorder of sleep disorders in which patients often suffer from

cognitive impairment. Recent neurological imaging studies have shown that cognitive impairment in OSA

patients is closely related to the extensive brain regions with abnormal neural activity. However, it remains

unclear whether the topological properties of function networks in OSA patients have changed. Based on

resting-states functional magenetic resonance imaging (rs-fMRI) and graph-analysis methods, this study

explores the different orgnization in functional brain network between OSA patients and healthy volunteers.

The brain connectome of patients with OSA exhibited small-worldness, but existed significant statistical

difference compared with health controls. Besides, the betweenness centrality and degree centrality of right

dorsolateral superior frontal gyrus and right hippocampal gyrus were significantly different in OSA patient

group. The aberrant topological properties illustrated that the functional integration and segregation of brain

networks in patients with OSA were disrupted.

1 INTRODUCTION

Obstructive sleep apnea (OSA) is the most common

adult respiratory disease in which patients suffer from

frequent apnea or hypoventilation due to recurrent

complete or incomplete upper airway obstruction

during sleep (Zhang, 2012). The main manifestations

of OSA patients are snoring during sleep with apnea

and superficial breathing, recurrent hypoxemia,

hypercarbia and sleep architecture disorders at night.

This leads to daytime sleepiness, cardio-pulmonary

vascular complications and even multi-organ

damage, which seriously affects the quality of life and

life expectancy of patients (He, 2009). OSA is an

independent risk factor for a variety of systemic

diseases and may significantly affect cognitive

function in addition to increasing the incidence of

hypertension, diabetes, respiratory failure and

cardiovascular disease (He, 2009). Current studies

have shown that OSA can extensively impair

cognitive function, including attentional alertness,

executive ability, memory and motor coordination,

and severely affect patient outcomes and prognosis

(Verstraeten, 2007; Aloia, 2004; Decary, 2000).

Resting-state functional magnetic resonance

imaging (MRI) provides a non-invasive and effective

technique for studying brain function. Previous

studies based on fMRI data have found that some

functional measures of regions exhibited functional

abnormalities in patients of OSA, such as the regional

homogeneity (ReHo) in the right temporal, parietal

and frontal regions (Santarnecchi, 2013), the

decreased ALFF of regions in default mode network

(DMN) (Li, Ma, 2016; Li, Nie, 2016). And

researchers have found that all these abnormal

regional alterations may be correlated to cognitive

dysfunction.

However, brain is a complex information

processing system (Khazaie, 2017). Any neurological

activity can’t occur as a result of a single neuron or

brain region working alone, but rather a collection of

neurons or several brain regions acting in concert

with each other to transmit and integrate information.

The complex network-based graph theory analysis

has been widely applied to characterize such complex

systems. The so-called network is a mathematical

representation of a real complex system, defined as a

collection of nodes and edges. For a brain network,

Hou, A., Pang, X. and Zhang, Q.

Altered Small-World Topological Properties of Functional Brain Network in Patients with OSA.

DOI: 10.5220/0012012400003633

In Proceedings of the 4th International Conference on Biotechnology and Biomedicine (ICBB 2022), pages 27-30

ISBN: 978-989-758-637-8

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

the nodes can be brain regions or voxels, or even

neurons, while edges can be identified by anatomical

connectivity, functional connectivity depending on

the characteristics of the data set. Functional

connectivity indicates a correlation in time between

the neurophysiological activity of two points. Based

on graph theory, Chen found that the whole brain

network of OSA patients exhibited decreased

smallworld topological property (Chen, 2018).

However, Huang found that the smallworld property

of the OSA brain network was not significantly

different to HC’s, but OSA performed significantly

lower in Cp and higher in Lp (Huang, 2019). Hence,

it follows that the functional network organization of

OSA patients are not clear yet.

In this paper, we intend to explore the impact of

OSA disease on patients' functional brain networks

from a complex network perspective.

2 MATERIAL METHOD

2.1 Data Acquisition

The data of this experiment were divided into two

groups, the OSA patient group and the healthy

controls group (HC), which included 24 OSA patients

with typical indications of OSA disease and met the

diagnostic criteria for the disease in the relevant draft;

21 healthy volunteers were also recruited as the

control group.

In this study, MRI data were acquired using a GE

Signa HDx magnetic resonance scanner with a field

strength of 3.0 T. Scans and resting-state fMRI data

were acquired using Gradient Recalled Echo (GRE)

single excitation planar echo imaging (EPI) sequence

with the following parameters: TR=2000ms,

TE=30ms, FA=90°, FOV was 240*240mm², a 64*64

matrix was used, thickness =3mm, the layer spacing

(gap) was 1mm, a total of 38 layers were divided, and

180 time points were acquired in each volume.

2.2 Data Pre-Processing

The DPARSF (Data Processing Assistant for Resting-

State fMRI) software based on MATLAB was used

to preprocess the functional MRI data of both OSA

and HC related to the following steps.

The images of the first 10 time points were

excluded to avoid the potential noise and instability.

Slice timing was applied to correct this time

difference. Head movement correction was used to

avoid a slight head movement. The unified standard

spatial EPI template with a voxel size of 3×3×3mm3

was used for transformation in order to facilitate the

later study. The whole brain average signal,

cerebrospinal fluid and white matter signal and

motion signal were regressed out. The linear drift was

removed. Then a band-pass filtering (0.01-0.08Hz)

was used to remove the interference of low frequency

and high frequency signals.

2.3 Construction of Brain Networks

and Calculation of Topological

Properties

Brain networks were constructed and topological

properties were calculated. Based on the human

Brainnetome Atlas template (Fan, 2016), the

cerebrum of each subject was divided into 246 brain

regions (nodes). The Pearson correlation coefficient

of the average time series of each two brain regions

was calculated as the functional connectivity (edges),

then followed by the Fisher r-to-z transformation to

normalize it. The z-scored 246×246 correlation

coefficient matrix was obtained.

The correlation matrices were binarized by a pre-

selected value of sparsity K (0.05 ≤ K ≤ 0.5). For a

specific K, we got an undirected binarized network

for each subject, then we applied graph theory to

calculate the topological properties of each brain

network. The global network properties contain

clustering coefficient (C

), characteristic path length

), normalized clustering coefficient (γ),

normalized characteristic path length (λ), and small-

world property (σ). The nodal properties contain the

betweenness centrality (BC) and degree centrality

(DC) of each brain regions.

2.4 Statistical Analysis

Two-samples t-test was performed for each parameter

corresponding to the two groups of subjects. p<0.05

is considered to be statistically different.

3 RESULTS AND DISCUSSIONS

3.1 Global Properties

The area under curve (AUC) of each global parameter

, L

, γ,λ, and σ) in OSA group and HC group is

shown in Figure 1(a). The Cp, Lp,λ of OSA patients

were significantly lower than healthy controls (p <

0.05), while theγand σ of OSA patients were

significantly higher than healthy control (p < 0.05).

ICBB 2022 - International Conference on Biotechnology and Biomedicine

In each sparsity, the small-world property σ was

greater than 1, which indicated that the brain

functional network in both the HC and OSA groups

had the small-world property. But the small

worldness of OSA patients was significantly

increased compared with healthy controls. These

results suggest that the small-world properties of the

functional brain network of OSA patients are

significantly altered.

(a) The difference of global topological properties

(b) Regions whose DC and BC showed between OSA and HC significantly different

Note: * p<0.05, which was considered significantly difference.

Figure 1: The AUC of global properties and local properties comparisons with OSA and HC.

3.2 Local Properties

In this paper, we mainly analyzed the local properties:

BC and DC of every brain region, and the regions

significantly different between OSA patients and

healthy controls are shown in Fig. 1(b) and Table 1.

The BC and DC of right dorsolateral superior frontal

gyrus (Frontal_Sup_R) was significantly lower in the

OSA patient group. The BC and DC of right

hippocampal gyrus (ParaHippocampal_R) was

significantly increased in OSA group.

Table 1: The P-value of brain regions whose BC and DC

were significantly different between OSA and HC..

Brain Regions BC DC

Frontal_Sup_R 0.0074 0.0048

ParaHippocampal_R 0.0010 0.0015

Altered Small-World Topological Properties of Functional Brain Network in Patients with OSA

4 CONCLUSION

Based on the results and discussions presented above,

the conclusions are obtained as below:

(1) The resting-state brain functional networks of

both OSA patients and normal controls exhibit small-

world property. But the small-world property of OSA

patients are significantly altered, which suggest that

the functional network organization of OSA patients

is altered.

(2) The local property of OSA patients was

significantly decreased in superior frontal gyrus,

which indicated that the impairment of the superior

frontal gyrus is associated with behavioral cognitive

dysfunction in OSA patients.

(3) The local property of OSA patients was

significantly increased in parahippocampal gyrus,

which indicated that the significantly higher value of

BC and DC may be related to compensatory

mechanisms of memory function impairment.

REFERENCES

Aloia MS, Amedt JT, Davis JD, et al. Neuropsychological

sequelae of obstructive sleep apnea-hypopnea

syndrome: a critical review [J]. J Int Neuropsychol Soc.

2004, 10(5): 772-785.

Chen L, Fan X, Li H, et al. Topological Reorganization of

the Default Mode Network in Severe Male Obstructive

Sleep Apnea[J]. Frontiers in neurology 2018; 9: 363.

Decary A, Rouleau 1, Montplaisir J. Cognitive deficits

associated with sleep apnea syndrome: a proposed

neuropsychological test battery [J]. Sleep. 2000, 23(3):

369-381.

Fan L, Li H, Zhuo J, et al. The Human Brainnetome Atlas:

A New Brain Atlas Based on Connectional Architecture

[J]. Cerebral Cortex, 2016, 26(8):3508-3526.

He Q Y, Chen B Y. Sleep and respiratory disease [M].

People's Medical Publishing House.2009; 89-421.

Huang Y, Liu Y, Zhao D, et al. Small-world properties of

the whole-brain functional networks in patients with

obstructive sleep apnea-hypopnea syndrome[J]. Sleep

medicine 2019; 62: 53-8.

Khazaie H, Veronese M , Noori K, et al. Functional

Reorganization in Obstructive Sleep Apnoea and

Insomnia: A Systematic Review of the Resting-State

fMRI[J]. Neuroscience & Biobehavioral Reviews,

2017, 77:219-231.

Li, C., Ma, X., Dong, M., Yin, Y., Hua, K., Li, M., Li, C.,

Zhan, W., Li, C., Jiang, G., 2016a. Abnormal

spontaneous regional brain activity in primary

insomnia: a resting-state functional magnetic resonance

imaging study. Neuropsychiatr. Dis. Treat. 12, 1371–

1378.

Li, H.J., Nie, X., Gong, H.H., Zhang, W., Nie, S., Peng,

D.C., 2016b. Abnormal resting-state functional

connectivity within the default mode network

subregions in male patients with obstructive sleep

apnea. Neuropsychiatr. Dis.Treat. 12, 203–212.

Santarnecchi, E., Sicilia, I., Richiardi, J., Vatti, G.,

Polizzotto, N.R., Marino, D., Rocchi, R., Van De Ville,

D., Rossi, A., 2013. Altered cortical and subcortical

local coherence in obstructive sleep apnea: a functional

magnetic resonance imaging study. J. Sleep Res. 22,

337–347.

Verstraeten E. Neurocognitive effects of obstructive sleep

apnea syndrome [J]. Curr Neurol Neurosci Rep. 2007,

7(2): 161-166.

Zhang Q. Resting-sate function magnetic resonance

imaging of obstructive sleep apnea-hypopnea syndrome

[D]. Tianjin Medical University, 2012.

ICBB 2022 - International Conference on Biotechnology and Biomedicine