A study on the alterations of negative functional connectivity and its diagnostic value in schizophrenia
10.3969/j.issn.1002-0152.2023.12.004
- VernacularTitle:精神分裂症患者脑功能负连接异常特征及其临床辅助诊断价值
- Author:
Shuxian YANG
1
,
2
;
Yingchao SONG
;
Yifan LI
;
Yayuan CHEN
;
Wen QIN
;
Meng LIANG
Author Information
1. 天津医科大学医学技术学院、医学影像学院(天津 300203)
2. 天津市功能影像重点实验室
- Keywords:
Schizophrenia;
Magnetic resonance imaging;
Resting-state functional magnetic resonance imaging;
Functional connectivity;
Clinical variables;
Multivariate pattern analysis;
Machine learning
- From:
Chinese Journal of Nervous and Mental Diseases
2023;49(12):727-733
- CountryChina
- Language:Chinese
-
Abstract:
Objective To assess the alterations of negative functional connectivity(FC),its relationship with clinical symptoms,and its potential value in schizophrenia(SZ).Methods Resting-state functional magnetic resonance imaging(rs-fMRI)data were acquired from patients with SZ and healthy controls(HC).For each participant,the whole brain image was first divided into 272 regions and then the FC between each pair of these regions was calculated using Pearson's correlation coefficient.Group-level negative FCs were identified using permutation test for each group.Each of the identified negative FCs was then compared between patients and controls to identify the altered negative FCs.Then,Spearman rank correlation was adopted to examine the relationship between the altered negative FCs and clinical variables.Finally,to evaluate the diagnostic value of negative FC in SZ,a multivariate pattern analysis(MVPA)was performed to distinguish between SZ and HC based on negative FCs.Results Ninety-one patients with SZ and 91 HC were included in this study,and 207 negative FCs in total were identified.Among the identified 207 negative FCs,12(constituting 5.80％of the total 207 negative FCs)were significantly altered in SZ compared with HC(Bonferroni correction,P<0.05),of which 11 were significantly decreased(i.e.,closer to 0)in SZ.The correlation analyses identified 2 significant associations-one was between a negative FC and the total score of the psychotic symptoms rating scales-auditory hallucinations(r=-0.24,P=0.02)and the other was between a negative FC and the weighted total score of the scale for the assessment of thought,language,and communication(r=0.26,P=0.01).Furthermore,the model for distinguishing between SZ and HC based on negative FCs achieved a classification accuracy of 72.6％that was significantly higher than chance-level accuracy(permutation test,P<0.001).Conclusion Negative FCs are altered in patients with SZ.Given that negative FCs are associated with clinical symptoms,thus they may serve as an imaging biomarker for assisting the diagnosis of SZ.
