DreaMR: Diffusion-driven Counterfactual Explanation for Functional MRI
Hasan A. Bedel and Tolga Çukur∗ This study was supported in part by a TUBITAK BIDEB scholarship awarded to H. A. Bedel, by TUBA GEBIP 2015 and BAGEP 2017 fellowships, and by a TUBITAK 1001 Grant 121N029 awarded to T. Çukur (Corresponding author: Tolga Çukur).H. A. Bedel and T. Çukur are with the Department of Electrical and Electronics Engineering, and the National Magnetic Resonance Research Center (UMRAM), Bilkent University, Ankara, Turkey, 06800. (e-mails: T. Çukur is also with the Neuroscience Graduate Program Bilkent University, Ankara, Turkey, 06800.
Abstract
Deep learning analyses have offered sensitivity leaps in detection of cognitive states from functional MRI (fMRI) measurements across the brain. Yet, as deep models perform hierarchical nonlinear transformations on their input, interpreting the association between brain responses and cognitive states is challenging. Among common explanation approaches for deep fMRI classifiers, attribution methods show poor specificity and perturbation methods show limited plausibility. While counterfactual generation promises to address these limitations, previous methods use variational or adversarial priors that yield suboptimal sample fidelity. Here, we introduce the first diffusion-driven counterfactual method, DreaMR, to enable fMRI interpretation with high specificity, plausibility and fidelity. DreaMR performs diffusion-based resampling of an input fMRI sample to alter the decision of a downstream classifier, and then computes the minimal difference between the original and counterfactual samples for explanation. Unlike conventional diffusion methods, DreaMR leverages a novel fractional multi-phase-distilled diffusion prior to improve sampling efficiency without compromising fidelity, and i
中文速览
用扩散模型（diffusion model）来解释大脑功能核磁共振（fMRI）深度分类器，是一个尚无人涉足的难题：现有的反事实解释方法依赖变分自编码器或生成对抗网络，生成质量有限，而传统扩散模型虽然生成质量高，却因采样步骤繁多、fMRI数据维度极高而效率低下。DreaMR提出了一种全新的"分数多阶段蒸馏扩散先验"（FMD），把扩散过程切分为若干连续片段并在每段内做多阶段蒸馏，配合线性复杂度的Transformer架构来捕捉fMRI扫描中的长程时空上下文，从而在极少步骤内高保真地生成反事实样本，再以原始样本与反事实样本之差来定位对分类决策至关重要的脑区。在静息态fMRI性别识别和任务态fMRI认知状态识别两项实验中，DreaMR在样本保真度、特异性和生成效率上均显著优于现有反事实方法，为可信赖的深度脑成像模型解释提供了坚实工具。
原文 arXiv:2307.09547;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2307.09547v1