BrainG3N：用於可控3D腦部MRI生成的雙重用途標記器

摘要

三维（3D）脑部磁共振成像（MRI）在临床神经病学与神经肿瘤学中占据核心地位。生成模型能够增强代表性不足的队列、模拟疾病演化轨迹，并支持隐私保护的数据共享。潜在扩散技术已成为医学影像数据建模的主流方案，但其对分词器提出两项竞争性需求：编码器嵌入必须保留下游任务所需的临床信息，而解码器需重建解剖学上逼真的体素图像。现有的基于重建驱动的分词器通常以牺牲前者为代价来实现后者。为应对这一挑战，我们提出了一种基于全自动掩码自编码器（MAE）的三维脑部MRI潜在扩散分词器，将编码器与解码器解耦：冻结的3D MAE编码器生成富含临床信息的嵌入特征，而专用卷积神经网络（CNN）解码器则从这些嵌入的线性投影中重建体素。我们在来自18个公开队列的35,309个体积数据上预训练编码器，这些数据涵盖四种模态、十种疾病类别及200余个采集站点，并在两种场景中验证其双重效用。首先，在包含23项任务的线性探测基准测试中，该编码器在21项任务上超越或持平当前最优模型（如BrainIAC、BrainSegFounder和MedicalNet）。其次，基于这些临床信息嵌入训练的扩散变换器（DiT）条件生成模型，既支持跨六个变量的条件生成，也支持患者特异性的纵向预测。综上，我们建立了一个统一的3D脑部MRI嵌入空间，该空间既能支持下游临床任务，又能实现可控生成。

English

Three-dimensional (3D) brain MRI is central to clinical neurology and neuro-oncology, where generative models could augment under-represented cohorts, simulate disease trajectories, and support privacy-preserving data sharing. Latent diffusion has been the go-to solution for modeling imaging data, but it places two competing demands on the tokenizer: encoder embeddings must retain the clinical information that downstream tasks act on, and the decoder must reconstruct anatomically faithful volumes. Existing reconstruction-driven tokenizers achieve the second at the expense of the first. To address this, we introduce a fully volumetric masked-autoencoder (MAE) based tokenizer for 3D brain MRI latent diffusion, decoupling encoder and decoder: a frozen 3D MAE encoder produces clinically informative embeddings, while a dedicated CNN decoder reconstructs voxels from a linear projection of those embeddings. We pretrain the encoder on 35,309 volumes from 18 public cohorts spanning four modalities, ten disease categories, and 200+ acquisition sites, and demonstrate its dual utility in two settings. First, on a 23-task linear-probing benchmark, the encoder outperforms or matches SOTA models (i.e., BrainIAC, BrainSegFounder, and MedicalNet) on 21 of 23 tasks. Second, a conditional diffusion transformer (DiT) trained on these clinically informative embeddings supports both conditional generation across six variables and patient-specific longitudinal forecasting. Together these results establish a single 3D brain-MRI embedding space capable of both downstream clinical tasks and controllable generation.