StimFlow：通过刺激条件化流匹配从图像合成大脑活动
StimFlow: Synthesizing Brain Activity from Images via Stimulus-conditioned Flow Matching
AI总结：
StimFlow是一种基于刺激条件化流匹配的模型，可直接在原生体素空间合成fMRI，仅用单个样本即可达到与五次采样平均的基线相当的性能，还能高效实现跨被试适配。
摘要：
功能磁共振成像（fMRI）响应在不同试次间存在固有变异性，这促使人们从生成视角看待刺激到响应的映射关系。现有生成式编码器通常在压缩潜在空间中运行，且主要以语义特征为条件，而语义特征为早期视觉皮层提供的低级线索有限。我们提出StimFlow，该模型采用带线性路径的条件流匹配，直接在原生体素空间中合成fMRI。此设计使训练与推理分布对齐，仅用单个样本即可实现有竞争力的性能，减少对平均处理的依赖。为匹配视觉层级，StimFlow将结构DINOv2特征与语义CLIP特征相结合，二者在皮层处理流中被证明具有互补性。在单个样本条件下，StimFlow的体素级皮尔逊相关系数达0.371，与需五次采样取平均的基线模型性能相当。在体素空间中生成可直接进行感兴趣区域（ROI）分析与概念定位，无需解码器。此外，仅用一小时扫描数据，它就能恢复超过半数全数据的性能，支持高效的跨被试适配。
英文摘要：
The inherent variability of fMRI responses across trials encourages a generative view of the stimulus-to-response mapping. Prior generative encoders often operate in compressed latent spaces and condition predominantly on semantic features, which provide limited low-level cues for early visual cortex. We propose StimFlow, which synthesizes fMRI directly in native voxel space using conditional flow matching with a linear path. This design aligns training and inference distributions, enabling competitive performance with a single sample and reducing reliance on averaging. To match the visual hierarchy, StimFlow combines structural DINOv2 features with semantic CLIP features, which prove complementary along the cortical processing stream. With a single sample, StimFlow achieves a per-voxel Pearson correlation of 0.371, matching baselines that average five draws. Generation in voxel space allows direct ROI analysis and concept localization without a decoder. Furthermore, it recovers over half of full-data performance from just one hour of scans, supporting efficient cross-subject adaptation.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
