Multimodal brain network topology and enhanced computer-aided diagnosis in Parkinson’s Disease: a systematic review and meta-analysis

Chao Zuo, Wei Liu, Huan Lan, Li Chen, N. N. Li, Yuying Yan, Chunyan Luo, Graham J. Kemp, Su Lui, Xueling Suo, Q. GongNpj 2026 · DOI

Abstract

Parkinson's disease (PD) is increasingly recognized as a brain network-disconnection syndrome. However, there is little consistent evidence on multimodal global topological alterations and their diagnostic value. We systematically searched PubMed, Embase and Web of Science up to March 2025 for articles reporting brain network topology in PD, to which we applied a multilevel random-effects meta-analyses with robust variance estimation to account for statistical dependencies. Our case-control meta-analysis included 80 studies (42 fMRI, 25 dMRI, 10 EEG, 4 sMRI, 3 others) involving 3736 PD patients and 2384 healthy controls. Compared to controls, PD patients showed lower structural and functional network segregation, especially when cognitively impaired. Structural network integration was also lower in PD, such deficits appearing to correlate with disease progression. Drug and network construction strategies were identified as potential moderating factors. Our diagnostic meta-analysis of 10 studies yielded a pooled diagnostic odds ratio of 16.4 and a pooled area under the curve of 0.86, with better diagnostic performance observed in studies using combined network metrics. These results support the clinical relevance of topological metrics in PD as potential biomarkers for disease characterization, prognosis and patient stratification, and underscore the importance of methodological harmonization and prospective validation in future research.

Introduction

Parkinson’s Disease (PD) has traditionally been diagnosed through the clinical observation of motor symptoms. However, these symptoms typically emerge only after significant neurological damage has already occurred. To enable earlier intervention, researchers have turned to "psychoradiology" to identify objective biomarkers within the brain's connectome—the complex map of neural connections.

While PD is increasingly viewed as a "brain network-disconnection syndrome," individual studies have often yielded inconsistent results due to small sample sizes and variations in imaging modalities (e.g., fMRI vs. dMRI). This systematic review and meta-analysis addresses these discrepancies by synthesizing data from the global research landscape to identify robust topological changes and evaluate their potential for computer-aided diagnosis (CAD).

Methods

The authors conducted a comprehensive search of PubMed, Embase, and Web of Science for studies published up to March 2025. They utilized a multilevel random-effects meta-analysis with robust variance estimation (RVE) to account for statistical dependencies within studies.

The analysis focused on Graph Theoretical Analysis (GTA), which models the brain as a mathematical graph of nodes (regions) and edges (connections). Key metrics evaluated included:

  • Network Segregation: The brain's capacity for local, specialized information processing (e.g., clustering coefficients, modularity).
  • Network Integration: The efficiency of communication across distant regions (e.g., path length, global efficiency).
  • Diagnostic Accuracy: Sensitivity, specificity, and Area Under the Curve (AUC) for PD classification.

The final dataset comprised 80 studies involving 3,736 PD patients and 2,384 healthy controls, spanning functional MRI (42 studies), diffusion MRI (25 studies), EEG (10 studies), and structural MRI (4 studies).

Results

Topological Alterations

The case-control meta-analysis revealed that PD patients consistently exhibit lower network segregation (reduced clustering and local efficiency) across multiple modalities. Structural network integration was also significantly impaired, with deficits appearing to correlate with disease progression. Notably, structural networks (dMRI) showed more severe disruption than functional networks (fMRI), suggesting that while physical connections suffer irreversible damage, functional networks may maintain some level of compensatory reorganization.

Cognitive and Technical Moderators

Network segregation deficits were significantly more pronounced in PD patients with cognitive impairment, identifying the breakdown of local specialized processing as a key driver of cognitive decline. Technical factors also influenced results; for example, 3T MRI scanners detected larger reductions in network clustering compared to 1.5T scanners, highlighting the importance of high-field imaging in biomarker detection.

Meta-analytic summary including Orchard plots of network metrics and the SROC curve for diagnostic performance.

Diagnostic Performance

The diagnostic meta-analysis of 10 classification studies demonstrated strong potential for clinical application. The pooled results showed a pooled AUC of 0.86 (95% CI: 0.83–0.89), a sensitivity of 0.77, and a specificity of 0.81. The Diagnostic Odds Ratio (DOR) reached 16.4. Performance was significantly enhanced when GTA metrics were combined with other imaging features (a "GTA Plus" strategy), rather than used in isolation.

Discussion

This synthesis confirms that brain network topology provides a powerful lens for understanding PD pathology. The findings support the use of these metrics as biomarkers for disease characterization and patient stratification. However, several challenges remain. The high level of heterogeneity ($I^2 > 75\%$) between studies underscores the need for methodological harmonization, particularly regarding medication status and network construction strategies. Furthermore, while the diagnostic results are promising, future research must prioritize external validation on independent, multi-center datasets and longitudinal designs to track disease progression over time.

In conclusion, redefining PD as a network-disconnection syndrome offers a robust framework for developing automated diagnostic tools, provided the field moves toward standardized imaging protocols.

Summary updated

Abstract

Parkinson's disease (PD) is increasingly recognized as a brain network-disconnection syndrome. However, there is little consistent evidence on multimodal global topological alterations and their diagnostic value. We systematically searched PubMed, Embase and Web of Science up to March 2025 for articles reporting brain network topology in PD, to which we applied a multilevel random-effects meta-analyses with robust variance estimation to account for statistical dependencies. Our case-control meta-analysis included 80 studies (42 fMRI, 25 dMRI, 10 EEG, 4 sMRI, 3 others) involving 3736 PD patients and 2384 healthy controls. Compared to controls, PD patients showed lower structural and functional network segregation, especially when cognitively impaired. Structural network integration was also lower in PD, such deficits appearing to correlate with disease progression. Drug and network construction strategies were identified as potential moderating factors. Our diagnostic meta-analysis of 10 studies yielded a pooled diagnostic odds ratio of 16.4 and a pooled area under the curve of 0.86, with better diagnostic performance observed in studies using combined network metrics. These results support the clinical relevance of topological metrics in PD as potential biomarkers for disease characterization, prognosis and patient stratification, and underscore the importance of methodological harmonization and prospective validation in future research.
BibTeX
@inproceedings{zuo2026multimodal,
  author = {Chao Zuo and Wei Liu and Huan Lan and Li Chen and N. N. Li and Yuying Yan and Li Li and Chunyan Luo and Graham J. Kemp and Su Lui and Xueling Suo and Q. Gong},
  title = {{Multimodal brain network topology and enhanced computer-aided diagnosis in Parkinson’s Disease: a systematic review and meta-analysis}},
  booktitle = {Npj},
  year = {2026},
  doi = {10.1038/s41746-025-02301-x},
}
Figure