Healthcare 5.0 represents the next evolution in intelligent and interconnected healthcare systems, leveraging emerging technologies such as Artificial Intelligence (AI) and the Internet of Medical Things (IoMT) to enhance patient care and automation. While Intrusion Detection Systems (IDSs) are a critical component for securing these environments, the current literature lacks a systematic analysis that jointly evaluates the effectiveness of AI models, the suitability of datasets, and the role of Explainable Artificial Intelligence (XAI) in the Healthcare 5.0 landscape. To fill this gap, this survey provides a comprehensive review of IDSs for Healthcare 5.0, analyzing state-of-the-art approaches and available datasets. Furthermore, a practical case study is presented, demonstrating that the fusion of network and biomedical features significantly improves threat detection, with physiological signals proving crucial for identifying complex attacks like spoofing. The primary contribution is therefore an integrated analysis that bridges the gap between cybersecurity theory and clinical practice, offering a guide for researchers and practitioners aiming to develop more secure, transparent, and patient-centric systems.

As redes CAN (Controller Area Network) são amplamente usadas na indústria automotiva e frequentemente alvo de ataques cibernéticos. A detecção desses ataques via aprendizado de máquina (AM) depende da seleção adequada de características para garantir o desempenho do modelo de predição. Este artigo propõe o IWSHAP, um novo método de seleção de características que combina o algorítimo Iterative Wrapper Subset Selection (IWSS) com os valores SHAP (SHapley Additive exPlanations). O principal objetivo é maximizar a performance do modelo de AM em um tempo reduzido. Os resultados indicam que IWSHAP consegue reduzir o número de características em até 99,17% e o tempo de execução em 98,3% comparado ao baseline.

The Greedy Randomized Adaptive Search Procedure for Feature Selection (GRASP-FS) is a recently-proposed metaheuristic that optimizes the feature selection process for Intrusion Detection Systems (IDS) by combining exploration and refinement techniques for more assertive intrusion detection. However, GRASP-FS may be time and resource-consuming for large datasets. In this work, we propose GRASPQ-FS, an extended version of GRASP-FS using Priority Queues to reduce resource consumption and processing time. As an additional contribution, we provide a comprehensive analysis of the most suitable parameters for our GRASPQ-FS. Our results reveal that GRASPQ-FS can speed up feature selection up to 90% over GRASP-FS, without compromising F1-Score. Also, we observed that a priority queue with 50 solutions saved 50% in execution time while increasing the F1-Score by 4.5%.

Este trabalho apresenta a ferramenta IWSHAP, que combina o algoritmo Iterative Wrapper Subset Selection (IWSS) com valores SHAP (SHapley Additive exPlanations) para encontrar o melhor conjunto de características e maximizar o desempenho de algoritmos de aprendizado de máquina. Os resultados em um cenário de detecção de intrusões em redes veiculares indicam que a ferramenta é eficaz na redução de datasets por meio da seleção de características, alcançando taxas elevadas de redução (e.g., 90%) e mantendo altas métricas de desempenho do modelo (acima de 90%).