BFLIDS: Blockchain-Driven Federated Learning for Intrusion Detection in IoMT Networks The Internet of Medical Things (IoMT) has significantly advanced healthcare but introduced critical security challenges. Traditional security solutions struggle with the dynamic nature of IoMT systems. Machine learning (ML)-based Intrusion Detection Systems (IDS) are increasingly used, but centralized ML approaches pose privacy risks due to single points of failure. Federated Learning (FL) offers a decentralized solution, enabling model updates on end devices without sharing private data. This study introduces BFLIDS, a blockchain-empowered FL-based IDS for IoMT networks. The approach leverages blockchain to secure transaction records, FL for data privacy, IPFS for decentralized storage, and MongoDB for efficient data management. Ethereum smart contracts oversee system interactions. The FedAvg algorithm was modified with Kullback–Leibler divergence estimation and adaptive weight calculation to enhance model accuracy and robustness. For classification, an Adaptive Max Pooling-based CNN and a modified BiLSTM with attention and residual connections were implemented on Edge-IIoTSet and TON-IoT datasets. The proposed BFLIDS achieved high accuracy (97.43%, 96.02%, 98.21%, 97.42%) in FL scenarios, competitive with centralized methods. The system effectively detects intrusions, enhancing IoMT security and privacy. The study addresses challenges in FL-based IDS, including data heterogeneity, adversarial attacks, and resource constraints. Blockchain provides secure, transparent ledger for model updates. The proposed BFLIDS integrates advanced ML techniques for intrusion detection, ensuring efficient and secure data management. The system's contributions include a novel blockchain-based FL approach, performance evaluation against centralized ML methods, and experimental outcomes showing competitive effectiveness. The study highlights the potential of blockchain and FL in enhancing IoMT security and privacy.BFLIDS: Blockchain-Driven Federated Learning for Intrusion Detection in IoMT Networks The Internet of Medical Things (IoMT) has significantly advanced healthcare but introduced critical security challenges. Traditional security solutions struggle with the dynamic nature of IoMT systems. Machine learning (ML)-based Intrusion Detection Systems (IDS) are increasingly used, but centralized ML approaches pose privacy risks due to single points of failure. Federated Learning (FL) offers a decentralized solution, enabling model updates on end devices without sharing private data. This study introduces BFLIDS, a blockchain-empowered FL-based IDS for IoMT networks. The approach leverages blockchain to secure transaction records, FL for data privacy, IPFS for decentralized storage, and MongoDB for efficient data management. Ethereum smart contracts oversee system interactions. The FedAvg algorithm was modified with Kullback–Leibler divergence estimation and adaptive weight calculation to enhance model accuracy and robustness. For classification, an Adaptive Max Pooling-based CNN and a modified BiLSTM with attention and residual connections were implemented on Edge-IIoTSet and TON-IoT datasets. The proposed BFLIDS achieved high accuracy (97.43%, 96.02%, 98.21%, 97.42%) in FL scenarios, competitive with centralized methods. The system effectively detects intrusions, enhancing IoMT security and privacy. The study addresses challenges in FL-based IDS, including data heterogeneity, adversarial attacks, and resource constraints. Blockchain provides secure, transparent ledger for model updates. The proposed BFLIDS integrates advanced ML techniques for intrusion detection, ensuring efficient and secure data management. The system's contributions include a novel blockchain-based FL approach, performance evaluation against centralized ML methods, and experimental outcomes showing competitive effectiveness. The study highlights the potential of blockchain and FL in enhancing IoMT security and privacy.