BrainCSD: A Hierarchical Consistency-Driven MoE Foundation Model for Unified Connectome Synthesis and Multitask Brain Trait Prediction
Xiongri Shen, Jiaqi Wang, Yi Zhong, Zhenxi Song, Leilei Zhao, Liling Li, Yichen Wei, Lingyan Liang, Shuqiang Wang, Baiying Lei, Demao Deng, Zhiguo Zhang
功能和结构连接(FC / SC)是用于大脑分析的关键多模态生物标志物,但它们的临床效用受到昂贵的获取,复杂的预处理和频繁缺失模式的阻碍。 现有的基础模型要么处理单一模式,要么缺乏跨模式和跨尺度一致性的明确机制。 我们提出了BrainCSD,一个分层专家混合(MoE)基础模型,共同合成FC / SC生物标志物并支持下游解码任务(诊断和治疗和预测)。 BrainCSD具有三个神经解剖学接地组件:(1)通过对比一致性将规范网络(例如,DMN,FPN)的区域激活与全局图集对齐的ROI特异性MoE;(2)编码激活MOE,在fMRI / dMRI中模拟动态跨时间/梯度依赖;(3)网络感知改进MoE,在个体中强制执行结构先验和对称。 在完整和缺失模式设置下对数据集进行评估,BrainCSD实现了SOTA结果:MCI与MCI的准确率为95.6%。 没有FC的CN分类,低合成误差(FC RMSE:0.038;SC RMSE:0.006),大脑年龄预测(MAE:4.04年)和MMSE评分估计(MAE:1.72分)。 代码可在 href{https://github.com/SXR3015/BrainCSD}{BrainCSD} 查阅。
Functional and structural connectivity (FC/SC) are key multimodal biomarkers for brain analysis, yet their clinical utility is hindered by costly acquisition, complex preprocessing, and frequent missing modalities. Existing foundation models either process single modalities or lack explicit mechanisms for cross-modal and cross-scale consistency. We propose BrainCSD, a hierarchical mixture-of-experts (MoE) foundation model that jointly synthesizes FC/SC biomarkers and supports downstream decoding...
