报告详情
FaultXAI: A Controlled Fault-Injection Study of Explanation Drift in ECG Classifiers
编号：86
访问权限：仅限参会人
更新：2026-07-22 16:09:48 浏览：19次
Online
摘要
Explainable AI (XAI) methods are increasingly deployed to interpret deep learning models in clinical decision support systems. However, the stability of these explanations under realistic signal degradation remains poorly understood. We introduce FaultXAI, a systematic framework for quantifying explanation drift the divergence between explanations of clean and corrupted inputs under controlled fault injection. We evaluate four clinically relevant fault types (Gaussian noise, baseline wander, lead dropout, segment dropout) across two benchmark ECG datasets: PTB-XL (12-lead, 21,837 records; 2,163 test records) and MIT-BIH (2-lead, 10,452 beats). Using Integrated Gradients and multiple stability metrics, our 5-seed experiments reveal: (1) additive faults exhibit strong cross-dataset agreement (r > 0.98), suggesting model intrinsic behaviour; (2) structural faults show dataset dependent effects; and (3) explanation drift can occur substantially even when predictive performance remains high (ρ < 0.7 at accuracy > 0.85). Our framework establishes reproducible benchmarks for
关键词
Explainable AI, ECG classification, explanation drift, fault injection, robustness, deep learning, clinical decision support
报告人
Deepak Chamarthi
Research Scholar Nagarjuna University稿件作者
全部评论
重要日期
-
会议日期
07月30日
2026
至08月01日
2026
-
07月28日 2026
初稿截稿日期
-
08月03日 2026
注册截止日期
主办单位
The United Societies of Science
承办单位
Kongunadu College of Engineering and Technology
协办单位
IEEE Section
IEEE Madras Section
IEEE Madras Section
联系方式
- ICNCCom 2026
- ic******@usssociety.org
历届会议
-
2026年07月30日 印度 Trichy
2026 International Conference on Networks Computers and Communications
发表评论