Yu Ji, Zhuo-Er Dong, Yi-Chong Duan, Li-Li Yao, Xiao-Rong Wu
{"title":"孔源性视网膜脱离中白质连接体的功能拓扑结构被破坏:来自图论和机器学习的见解。","authors":"Yu Ji, Zhuo-Er Dong, Yi-Chong Duan, Li-Li Yao, Xiao-Rong Wu","doi":"10.1097/WNR.0000000000002190","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Rhegmatogenous retinal detachment (RRD) is known to induce functional alterations in the gray matter regions associated with vision. However, the impact of RRD on the white matter (WM) connectome remains largely unexplored.</p><p><strong>Methods: </strong>We applied graph theory to evaluate the functional network topology of the WM connectome in RRD patients. A support vector machine (SVM) classifier, combined with SHapley Additive exPlanations (SHAP), was then employed to distinguish RRD patients from healthy controls (HCs) and to identify key brain regions driving model predictions.</p><p><strong>Results: </strong>Compared to HCs, RRD patients exhibited significant disruptions in both global and nodal network topology. Network-based statistics identified 23 subnetworks with altered connectivity. Notably, the integration of SVM and SHAP analyses revealed that betweenness centrality (Bc) was the most discriminative topological feature, achieving an area under the curve of 0.9211.</p><p><strong>Conclusion: </strong>These findings suggest that RRD disrupts critical hubs within the central visual and higher-order cognitive networks, leading to characteristic network reorganization. Moreover, Bc shows promise as an early neuroimaging biomarker for RRD. Overall, our results advance the understanding of neuroadaptive changes in RRD and support the clinical application of network topological metrics in early diagnosis.</p>","PeriodicalId":19213,"journal":{"name":"Neuroreport","volume":" ","pages":""},"PeriodicalIF":1.6000,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Disrupted functional topology of the white matter connectome in rhegmatogenous retinal detachment: insights from graph theory and machine learning.\",\"authors\":\"Yu Ji, Zhuo-Er Dong, Yi-Chong Duan, Li-Li Yao, Xiao-Rong Wu\",\"doi\":\"10.1097/WNR.0000000000002190\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Background: </strong>Rhegmatogenous retinal detachment (RRD) is known to induce functional alterations in the gray matter regions associated with vision. However, the impact of RRD on the white matter (WM) connectome remains largely unexplored.</p><p><strong>Methods: </strong>We applied graph theory to evaluate the functional network topology of the WM connectome in RRD patients. A support vector machine (SVM) classifier, combined with SHapley Additive exPlanations (SHAP), was then employed to distinguish RRD patients from healthy controls (HCs) and to identify key brain regions driving model predictions.</p><p><strong>Results: </strong>Compared to HCs, RRD patients exhibited significant disruptions in both global and nodal network topology. Network-based statistics identified 23 subnetworks with altered connectivity. Notably, the integration of SVM and SHAP analyses revealed that betweenness centrality (Bc) was the most discriminative topological feature, achieving an area under the curve of 0.9211.</p><p><strong>Conclusion: </strong>These findings suggest that RRD disrupts critical hubs within the central visual and higher-order cognitive networks, leading to characteristic network reorganization. Moreover, Bc shows promise as an early neuroimaging biomarker for RRD. Overall, our results advance the understanding of neuroadaptive changes in RRD and support the clinical application of network topological metrics in early diagnosis.</p>\",\"PeriodicalId\":19213,\"journal\":{\"name\":\"Neuroreport\",\"volume\":\" \",\"pages\":\"\"},\"PeriodicalIF\":1.6000,\"publicationDate\":\"2025-06-24\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Neuroreport\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.1097/WNR.0000000000002190\",\"RegionNum\":4,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"NEUROSCIENCES\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Neuroreport","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1097/WNR.0000000000002190","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"NEUROSCIENCES","Score":null,"Total":0}
Disrupted functional topology of the white matter connectome in rhegmatogenous retinal detachment: insights from graph theory and machine learning.
Background: Rhegmatogenous retinal detachment (RRD) is known to induce functional alterations in the gray matter regions associated with vision. However, the impact of RRD on the white matter (WM) connectome remains largely unexplored.
Methods: We applied graph theory to evaluate the functional network topology of the WM connectome in RRD patients. A support vector machine (SVM) classifier, combined with SHapley Additive exPlanations (SHAP), was then employed to distinguish RRD patients from healthy controls (HCs) and to identify key brain regions driving model predictions.
Results: Compared to HCs, RRD patients exhibited significant disruptions in both global and nodal network topology. Network-based statistics identified 23 subnetworks with altered connectivity. Notably, the integration of SVM and SHAP analyses revealed that betweenness centrality (Bc) was the most discriminative topological feature, achieving an area under the curve of 0.9211.
Conclusion: These findings suggest that RRD disrupts critical hubs within the central visual and higher-order cognitive networks, leading to characteristic network reorganization. Moreover, Bc shows promise as an early neuroimaging biomarker for RRD. Overall, our results advance the understanding of neuroadaptive changes in RRD and support the clinical application of network topological metrics in early diagnosis.
期刊介绍:
NeuroReport is a channel for rapid communication of new findings in neuroscience. It is a forum for the publication of short but complete reports of important studies that require very fast publication. Papers are accepted on the basis of the novelty of their finding, on their significance for neuroscience and on a clear need for rapid publication. Preliminary communications are not suitable for the Journal. Submitted articles undergo a preliminary review by the editor. Some articles may be returned to authors without further consideration. Those being considered for publication will undergo further assessment and peer-review by the editors and those invited to do so from a reviewer pool.
The core interest of the Journal is on studies that cast light on how the brain (and the whole of the nervous system) works.
We aim to give authors a decision on their submission within 2-5 weeks, and all accepted articles appear in the next issue to press.