DSTAN-Med: Dual-Channel Spatiotemporal Attention with Physiological Plausibility Filtering for False Data Injection Attack Detection in IoT-Based Medical Devices

2026-05-13

AI Summary

This study addresses the critical challenge of false data injection (FDI) attacks in Internet-of-Medical-Things devices, which can manipulate vital signs and endanger patient safety. Existing approaches struggle to disentangle spatiotemporal anomalies effectively and often lack physiological plausibility constraints. To overcome these limitations, this work proposes DSTAN-Med, a novel framework that employs a dual-channel self-attention mechanism to separately model inter-sensor spatial correlations and temporal dependencies, augmented with a residual 1D-CNN for local temporal feature extraction. Notably, it introduces—for the first time—a parameter-free Physiological Plausibility Filter (PPF) to suppress attack signals that violate medical常识. Evaluated on PhysioNet/CinC 2012, MIMIC-III, and WESAD datasets, DSTAN-Med achieves an average sensitivity gain of 7.4–8.3 percentage points (p<0.01) over the strongest Transformer-based baseline, TranAD, with the PPF alone contributing 3.1–4.2 percentage points of independent accuracy improvement.

Abstract

False data injection (FDI) attacks on Internet of Medical Things (IoMT) sensor streams falsify vital signs in transit, threatening patient safety and defeating clinical monitoring systems that lack cyber-physical anomaly detection capability. Existing deep learning detectors conflate inter-sensor spatial correlations with temporal dependencies in a shared latent space, preventing disentanglement of the distinct spatial and temporal signatures that FDI attacks imprint simultaneously; no current method exploits domain knowledge to constrain outputs against physiologically impossible attack patterns. We propose DSTAN-Med, a supervised framework comprising a Dual-channel Attention Mechanism (DAM) that routes multivariate sensor windows through independent sensor-wise (SWA) and time-wise (TWA) self-attention pathways operating on orthogonal tensor axes, a residual 1D-CNN block for local temporal feature extraction, and a zero-parameter Physiological Plausibility Filter (PPF) that suppresses attack signatures violating domain-knowledge bounds. Evaluated across three IoMT sensor datasets - PhysioNet/CinC 2012 (ICU vital signs), MIMIC-III Waveform (continuous ICU waveforms), and WESAD (wearable biosensor signals) - DSTAN-Med achieves mean sensitivity gains of 7.4-8.3 percentage points over the strongest Transformer baseline (TranAD), with improvements significant at p < 0.01 (McNemar's test, Holm-Bonferroni correction). The PPF contributes independent precision gains of 3.1-4.2 percentage points at negligible sensitivity cost across all three corpora. Ablation studies confirm that each component is individually necessary; removal of residual connections alone reduces sensitivity by 14.0 percentage points. The source code is publicly available at https://github.com/mehedi93hasan/DSTAN-MED.

Problem

• False Data Injection
• Internet of Medical Things
• Anomaly Detection
• Physiological Plausibility
• Cyber-Physical Security

Innovation

• Dual-channel Attention Mechanism
• Physiological Plausibility Filter
• False Data Injection Detection
• IoMT Security
• Spatiotemporal Disentanglement