Dupré la Tour Tom, Visconti di Oleggio Castello Matteo, Gallant Jack L
Helen Wills Neuroscience Institute, University of California, Berkeley, CA, United States.
Department of Neuroscience, University of California, Berkeley, CA, United States.
Imaging Neurosci (Camb). 2025 May 9;3. doi: 10.1162/imag_a_00575. eCollection 2025.
The Voxelwise Encoding Model framework (VEM) is a powerful approach for functional brain mapping. In the VEM framework, features are extracted from the stimulus (or task) and used in an encoding model to predict brain activity. If the encoding model is able to predict brain activity in some part of the brain, then one may conclude that some information represented in the features is also encoded in the brain. In VEM, a separate encoding model is fitted on each spatial sample (i.e., each voxel). VEM has many benefits compared to other methods for analyzing and modeling neuroimaging data. Most importantly, VEM can use large numbers of features simultaneously, which enables the analysis of complex naturalistic stimuli and tasks. Therefore, VEM can produce high-dimensional functional maps that reflect the selectivity of each voxel to large numbers of features. Moreover, because model performance is estimated on a separate test dataset not used during fitting, VEM minimizes overfitting and inflated Type I error confounds that plague other approaches, and the results of VEM generalize to new subjects and new stimuli. Despite these benefits, VEM is still not widely used in neuroimaging, partly because no tutorials on this method are available currently. To demystify the VEM framework and ease its dissemination, this paper presents a series of hands-on tutorials accessible to novice practitioners. The VEM tutorials are based on free open-source tools and public datasets, and reproduce the analysis presented in previously published work.
体素编码模型框架（VEM）是一种用于功能性脑图谱绘制的强大方法。在VEM框架中，从刺激（或任务）中提取特征，并将其用于编码模型以预测脑活动。如果编码模型能够预测大脑某些部分的脑活动，那么可以得出结论，特征中所表示的某些信息也在大脑中进行了编码。在VEM中，在每个空间样本（即每个体素）上拟合一个单独的编码模型。与其他分析和建模神经影像数据的方法相比，VEM有许多优点。最重要的是，VEM可以同时使用大量特征，这使得能够分析复杂的自然主义刺激和任务。因此，VEM可以生成反映每个体素对大量特征选择性的高维功能图谱。此外，由于模型性能是在拟合过程中未使用的单独测试数据集上进行估计的，VEM将困扰其他方法的过拟合和膨胀的I型错误混淆降至最低，并且VEM的结果可以推广到新的受试者和新的刺激。尽管有这些优点，VEM在神经影像领域仍然没有得到广泛应用，部分原因是目前尚无关于该方法的教程。为了揭开VEM框架的神秘面纱并促进其传播，本文提供了一系列新手从业者可以上手的教程。VEM教程基于免费的开源工具和公共数据集，并重现了先前发表作品中所呈现的分析。
