When Language Models Meet NeuroGraphs: Exploring Enhanced Agentic LLM Framework Towards Brain Network Analysis

2026-07-24

AI Summary

This work addresses the limitations of existing deep learning approaches in brain network analysis, which often treat the task as a black-box classification problem lacking interpretability, and the shortcomings of general-purpose large language models (LLMs), which suffer from a structural–linguistic gap, insufficient neuroscientific knowledge, and overconfidence. To overcome these challenges, the authors propose BrainAgent—the first agent-based LLM framework specifically designed for brain network analysis—reformulating connectome classification as an iterative process of topology-aware description, knowledge retrieval, reasoning, and reflection. Integrating multi-level graph structural encoding, neuroscience-informed knowledge augmentation, and verifiable reasoning mechanisms, BrainAgent significantly outperforms current baselines across four resting-state fMRI datasets and demonstrates consistent improvements regardless of whether open- or closed-source LLM backbones are used, yielding multi-granular, interpretable, and verifiable analytical outcomes.

Abstract

Brain network analysis is crucial for understanding cognition and neurological disorders, yet existing deep learning methods mainly treat connectome analysis as a graph-to-logit classification problem, offering limited explanatory reasoning. Large language models (LLMs) provide a promising interface for knowledge-intensive scientific analysis, but directly applying general-purpose LLMs to brain networks remains challenging due to the structure-language gap, limited neuroscience grounding, and overconfident positive predictions. In this paper, we propose \textbf{BrainAgent}, an agentic LLM framework for knowledge-enhanced brain network analysis. BrainAgent reformulates connectome classification as an iterative process of topology-aware understanding, external retrieval, reasoning, and reflection. Specifically, it first converts raw brain networks into compact multi-level structural descriptions through brain-specific analysis tools, then retrieves relevant neuroscience knowledge and task-specific cases to ground the reasoning process, and finally generates structured predictions with reflective verification. Experiments on four public rs-fMRI datasets show that BrainAgent consistently improves different closed-source and open-source LLM backbones over direct prompting and standard reasoning baselines. Further ablation and interpretability analyses demonstrate the effectiveness of each component and show that BrainAgent produces more comprehensive, multi-level, and verifiable explanations.These results indicate that agentic LLMs provide a practical route toward interpretable and knowledge-grounded brain network analysis.

Problem

• brain network analysis
• connectome classification
• large language models
• interpretability
• neuroscience grounding

Innovation

• Agentic LLM
• Brain Network Analysis
• Knowledge Retrieval
• Topology-aware Reasoning
• Interpretable AI