发表机构
College of Medicine and Biological Information Engineering, Northeastern University（东北大学医学与生物信息工程学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究在虚拟现实平衡任务中检验伪迹去除对皮肤电活动分类的影响，发现门控网络虽改善波形但未提升分类性能，强调预处理应基于下游决策评估。
AI 中文摘要
伪迹去除通常在皮肤电活动（EDA）分类之前进行，其假设是更干净的信号能支持更好的决策。我们在一个虚拟现实（VR）平衡扰动任务中检验了这一假设。一个残差门控网络在带有专家校正EDA的基准数据集上训练，冻结后应用于VR记录，其中原始信号和门控信号由五种已发表的时间序列方法在相同的留一参与者评估下进行分类。在基准数据集上，门控能很好地检测伪迹（记录中位数AUROC为0.94），并将伪迹区域内的误差降低17.8%。在VR任务中，它并未改善分类。平衡准确率的变化范围从-1.35到+0.93个百分点，没有分类器得到改善，两个分类器损失了准确率，所有五个分类器与原始输入在±3.32个百分点内等效。收益在波形和决策之间丢失了。降低波形误差的校正也减少了所有43个基准记录中的皮肤电导反应检测。处理使92.8%的预测保持不变，且其改变的预测中，被纠正和被破坏的比例相近。VR记录也携带很少的污染（估计4.6%的样本），即使对故意注入伪迹的完美定位，在最敏感的分类器中也仅恢复3.3个百分点。伪迹水平与准确率之间的汇总关联（11.3个百分点）在参与者内部消失（0.1个百分点），表明个体差异如何使清洗看起来有用。预处理应根据其支持的决策来评判，并与未处理的对照组进行比较。
英文摘要
Artifact removal routinely precedes the classification of electrodermal activity (EDA), on the assumption that a cleaner signal supports a better decision. We tested this assumption in a virtual-reality (VR) balance-disturbance task. A residual gating network was trained on a benchmark with expert-corrected EDA, frozen, and applied to VR recordings, where raw and gated signals were classified by five published time-series methods under identical leave-one-participant-out evaluation. On the benchmark the gate detected artifacts well (median record AUROC 0.94) and reduced error inside artifact regions by 17.8%. In the VR task it did not improve classification. Changes in balanced accuracy ranged from -1.35 to +0.93 percentage points, no classifier improved and two lost accuracy, and all five were equivalent to raw input within +/- 3.32 points. The benefit was lost between waveform and decision. The correction that lowered waveform error also reduced skin conductance response detection in all 43 benchmark records. Processing left 92.8% of predictions unchanged, and the predictions it did change were corrected and corrupted at similar rates. The VR recordings also carried little contamination (an estimated 4.6% of samples), and even perfect localization of deliberately injected artifacts recovered only 3.3 points in the most sensitive classifier. A pooled association between artifact level and accuracy (11.3 points) disappeared within participants (0.1 points), showing how differences between people can make cleaning look useful. Preprocessing should be judged by the decision it supports, against an unprocessed arm.
Comments10 pages, 9 figures, 3 tables. Code and frozen data: https://github.com/rtb-1005/FairEDA
