Accurate delineation of tumor-adjacent functional brain regions is essential for planning function-preserving neurosurgery. Functional magnetic resonance imaging (fMRI) is increasingly used for presurgical counseling and planning. When analyzing presurgical fMRI data, false negatives are more dangerous to the patients than false positives because patients are more likely to experience significant harm from failing to identify functional regions and subsequently resecting critical tissues. In this paper, we propose a novel spatially adaptive variable screening procedure to enable effective control of false negatives while leveraging the spatial structure of fMRI data. Compared to existing statistical methods in fMRI data analysis, the new procedure directly controls false negatives at a desirable level and is completely data-driven. The new method is also substantially different from existing false-negative control procedures which do not take spatial information into account. Numerical examples show that the new method outperforms several state-of-the-art methods in retaining signal voxels, especially the subtle ones at the boundaries of functional regions, while providing cleaner separation of functional regions from background noise. Such results could be valuable to preserve critical tissues in neurosurgery.

翻译：精确描绘肿瘤附近的功能性脑区对于规划保留功能的神经外科手术至关重要。功能性磁共振成像（fMRI）越来越多地被用于术前咨询和手术规划。在分析术前fMRI数据时，假阴性对患者的危害大于假阳性，因为未能识别功能区并随后切除关键组织更可能导致患者遭受显著损伤。本文提出了一种新颖的空间自适应变量筛选程序，能够在利用fMRI数据空间结构的同时有效控制假阴性。与现有fMRI数据分析中的统计方法相比，该新程序能直接以理想水平控制假阴性，且完全由数据驱动。该方法也显著区别于现有不考虑空间信息的假阴性控制程序。数值示例表明，新方法在保留信号体素方面优于多种最新方法，尤其是功能区域边界处的细微信号，同时能将功能区域与背景噪声更清晰地分离。此类结果对神经外科手术中保护关键组织具有重要价值。