MedDefender-MHAN: Healthcare IoT Intrusion Detection Benchmark Results

Ali Alqazzaz published this dataset on 2026-04-17. It contains results for the MedDefender-MHAN explainable multi-head attention network, an intrusion detection system designed for healthcare Internet of Things (IoMT) environments. The data likely includes performance metrics from evaluations on the CICIDS2017 and TON_IoT benchmark datasets.

Use Cases

• Benchmarking intrusion detection models based on reported detection accuracies of 99.47% and 98.92%.
• Evaluating explainable AI methods in cybersecurity based on the embedded interpretability mechanism described.
• Analyzing model inference performance based on the sub-3ms latency and 435 samples per second throughput metrics.
• Comparing attack signature alignment based on the 94.6% alignment with expert annotations mentioned in the description.

Strengths

• Reported high detection accuracies of 99.47% and 98.92% on benchmark datasets.
• Includes specific performance metrics: sub-3ms inference latency and throughput of 435 samples per second.
• Explainability metrics are provided, with 94.6% alignment and 91.9% temporal accuracy.
• Released under a permissive CC-BY-4.0 license.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is very small at 5.5 KB, indicating limited scope, likely containing summary results rather than raw data.

Provenance

Source

figshare, author Ali Alqazzaz.

Collection Method

Likely contains results from evaluating the MedDefender-MHAN framework on benchmark datasets.

Time Range

null

Freshness

Last updated 2026-04-17 17:39:42; freshness should be verified.

Geography

null