Recent years have seen a growing interest in methods for predicting a variable of interest, such as a subject's diagnosis, from medical images. Methods based on discriminative modeling excel at making accurate predictions, but are challenged in their ability to explain their decisions in anatomically meaningful terms. In this paper, we propose a simple technique for single-subject prediction that is inherently interpretable. It augments the generative models used in classical human brain mapping techniques, in which cause-effect relations can be encoded, with a multivariate noise model that captures dominant spatial correlations. Experiments demonstrate that the resulting model can be efficiently inverted to make accurate subject-level predictions, while at the same time offering intuitive causal explanations of its inner workings. The method is easy to use: training is fast for typical training set sizes, and only a single hyperparameter needs to be set by the user. Our code is available at https://github.com/chiara-mauri/Interpretable-subject-level-prediction.

翻译：近年来，基于医学图像预测感兴趣变量（如受试者诊断结果）的方法日益受到关注。基于判别式建模的方法在做出准确预测方面表现出色，但在以解剖学上有意义的术语解释其决策方面面临挑战。本文提出了一种简单且本质上可解释的个体预测技术。该方法在经典人脑映射技术所使用的生成式模型（其中可编码因果关系）基础上，引入了一个捕获主要空间相关性的多元噪声模型。实验表明，所得模型能够高效反转以做出准确的个体级预测，同时提供其内部工作机制的直观因果解释。该方法易于使用：在典型训练集规模下训练速度快，且用户仅需设置一个超参数。我们的代码可在 https://github.com/chiara-mauri/Interpretable-subject-level-prediction 获取。