{"paper":{"title":"DSTAN-Med: Dual-Channel Spatiotemporal Attention with Physiological Plausibility Filtering for False Data Injection Attack Detection in IoT-Based Medical Devices","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A dual-channel attention model plus a zero-parameter plausibility filter detects falsified vital signs on IoMT sensors with 7.4-8.3 point sensitivity gains over Transformer baselines.","cross_cats":[],"primary_cat":"cs.CR","authors_text":"Md Mehedi Hasan, Md Zakir Hossain, Rafiqul Islam","submitted_at":"2026-05-13T22:39:47Z","abstract_excerpt":"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 patte"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"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).","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the zero-parameter Physiological Plausibility Filter will not suppress genuine but unusual physiological states that still fall inside the chosen bounds, and that the synthetic or injected attack patterns used in the three evaluation corpora match the distribution of real-world FDI attacks.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"DSTAN-Med uses separate sensor-wise and time-wise attention plus a zero-parameter physiological filter to detect falsified vital signs, reporting 7.4-8.3 percentage point sensitivity gains over Transformer baselines on three public datasets.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A dual-channel attention model plus a zero-parameter plausibility filter detects falsified vital signs on IoMT sensors with 7.4-8.3 point sensitivity gains over Transformer baselines.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"d7ed18496727cf0936dacf42f3ec416dceb3a4cd12f1a8b1195dcb1b11f58771"},"source":{"id":"2605.14165","kind":"arxiv","version":1},"verdict":{"id":"9d9b562f-74a8-4e78-8029-000dd9827bd7","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T04:53:43.260585Z","strongest_claim":"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).","one_line_summary":"DSTAN-Med uses separate sensor-wise and time-wise attention plus a zero-parameter physiological filter to detect falsified vital signs, reporting 7.4-8.3 percentage point sensitivity gains over Transformer baselines on three public datasets.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the zero-parameter Physiological Plausibility Filter will not suppress genuine but unusual physiological states that still fall inside the chosen bounds, and that the synthetic or injected attack patterns used in the three evaluation corpora match the distribution of real-world FDI attacks.","pith_extraction_headline":"A dual-channel attention model plus a zero-parameter plausibility filter detects falsified vital signs on IoMT sensors with 7.4-8.3 point sensitivity gains over Transformer baselines."},"references":{"count":55,"sample":[{"doi":"","year":2025,"title":"A deep reinforcement learning-based robust intrusion detection system for securing IoMT healthcare networks,","work_id":"2d6b45bb-bf16-4bdc-ac26-b1dba58ff870","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Implementing anomaly-based intrusion detection for resource-constrained devices in IoMT networks,","work_id":"876b8e2d-8db2-4131-944a-8ce00e155f2e","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Internet of medical things (IoMT) market size, share & industry analysis,","work_id":"38694570-72e0-4370-8a32-a642f5b54294","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"A novel internet of medical things hybrid model for cybersecurity anomaly detection,","work_id":"7fb551b6-ea5f-49a2-96d2-2c444114c493","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Unpatched and outdated medical devices provide cyber attack opportunities,","work_id":"89039a6a-a408-4a4d-a399-cf08c9fbae84","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":55,"snapshot_sha256":"a8a204b083395eeae82e4fe3c4ffbf6aa765d8f42194db89bb22db4021568749","internal_anchors":1},"formal_canon":{"evidence_count":2,"snapshot_sha256":"1c4c40dfa981c402f91664abdbfa05b3f25171a59553e8bc6c0d3b6836e3e3af"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}