Related Experiment Video
Updated: Jul 1, 2026
05:30
Soft Pneumatic Robot Modulates Graph Theory Metrics of Brain Network for Hand Rehabilitation After Stroke
Published on: October 10, 2025
Personalizing post-stroke symptom management: integrating network analysis and in silico intervention to identify
Yiqing Zhang1, Aizhen Wang2, Qun Fang1
1Department of Nursing, Ningbo Medical Center Lihuili Hospital, Ningbo, 315000, China.
BMC Neurology
|June 29, 2026
Summary
Stroke survivors exhibit diverse symptom experiences, falling into low and high burden profiles with distinct network structures. Targeting central symptoms offers the greatest burden reduction, but worsening non-central symptoms can significantly increase overall burden.
Area of Science:
- Neurology
- Psychology
- Computational Modeling
Background:
- Stroke is a leading cause of death and disability globally.
- Survivors experience complex, interacting symptoms, leading to significant burden and reduced quality of life.
- Individual symptom experiences in stroke survivors are highly heterogeneous, necessitating personalized management strategies.
Purpose of the Study:
- Identify distinct symptom burden profiles in stroke survivors.
- Compare symptom network structures across different patient profiles.
- Simulate potential intervention targets using computational modeling to guide symptom management.
Main Methods:
- Recruited 451 stroke survivors from Zhejiang Province, China.
- Utilized Stroke Symptom Experience Scale, exploratory factor analysis, and latent profile analysis to identify symptom domains and patient profiles.
- Employed network analysis (EBICglasso) and NodeIdentifyR for network estimation, comparison, and intervention simulation.
Main Results:
- Identified four symptom domains: Motor Dysfunction, Emotional Disorders, Cognitive and Language Disorders, and Pain/Foot Abnormalities.
- Revealed two distinct profiles: low symptom burden (65.4%) and high symptom burden (34.6%).
- Central symptoms varied by profile; targeting high-centrality symptoms predicted significant burden reduction, while worsening specific non-central symptoms (e.g., foot varus) in the low-burden group increased burden substantially.
Conclusions:
- Stroke survivor symptom experience is heterogeneous, classifiable into distinct burden profiles with unique network structures.
- Computational modeling suggests a benefit-risk framework for symptom management, highlighting the importance of targeting high-centrality symptoms while considering potential risks of interventions on non-central symptoms.
- Findings support hypothesis generation for tailored symptom management, emphasizing the need for future longitudinal validation.
