Journal of Electronic & Information Systems

An Intelligent Clinical Decision Support Framework for Heart Disease Risk Prediction

2026-05-21T14:46:08+08:00

Cardiovascular diseases (CVDs) are among the leading causes of mortality worldwide. Early detection and risk stratification are critical for preventive care. Traditional machine learning (ML) models can predict heart disease risk but often lack interpretability and fail to integrate with real-time clinical data. Recent advances in fine-tuned large language models (Custom GPT) offer natural language explanations but are limited by insufficient interoperability with heterogeneous healthcare data sources. This study aims to design and evaluate a Model Context Protocol (MCP)–enabled Custom GPT framework that integrates ML-based predictive models with external healthcare systems—including EHRs, laboratory APIs, and wearable devices—to deliver context-aware, explainable, and clinically actionable heart disease risk predictions. The experimental evaluation was conducted on a validated cardiovascular dataset containing 303 patient records and 14 clinically relevant attributes derived from publicly available clinical repositories. Experimental evaluation demonstrated improved predictive accuracy (approximately 88% with the XGBoost ensemble) and robustness compared to standalone models. MCP integration enabled dynamic contextual awareness, reduced latency in tool orchestration, and enriched interpretability through RAG-based explanations. Clinician and patient evaluations confirmed enhanced usability and transparency. This approach paves the way for broader adoption of agentic AI in clinical workflows.

A Hybrid Artificial Intelligence Approach for Intrusion Detection in Wireless Sensor Networks

2026-05-28T15:57:08+08:00

Wireless sensor networks (WSNs) play a vital role in modern applications such as environmental monitoring, industrial automation, and smart infrastructure, where reliable data transmission, robustness, and energy efficiency are essential. However, their distributed architecture and constrained computational resources make them highly vulnerable to a wide range of security threats, including Blackhole, Grayhole, Flooding, and Scheduling attacks. These attacks can severely disrupt network functionality, degrade data integrity, and compromise the overall reliability of mission-critical operations. To address these challenges, we present an intrusion detection system (IDS) framework that leverages a diverse set of machine learning (ML) models, incorporating both boosting and non-boosting techniques, as well as deep learning (DL) architectures, including sequential and non-sequential designs. This diversity enables the framework to capture varied learning behaviors and decision boundaries. To further enhance detection accuracy and adaptability, Ant colony optimization (ACO) is employed as a metaheuristic tuning layer, refining hyperparameters to improve performance under the strict resource limitations typical of WSN environments. Each model is evaluated in both its baseline and ACO-optimized form, enabling a detailed comparative analysis that highlights the influence of optimization on intrusion detection effectiveness. Experimental results demonstrate that ACO significantly strengthens model resilience against diverse threats, offering an adaptive and efficient approach to securing modern WSN infrastructures.

Enhancing Software Defect Prediction Security via Federated BLSTM with Differential Privacy

2026-06-01T09:53:27+08:00

Software Defect Prediction (SDP) is an issue of paramount importance for improvement of software quality but data isolation and privacy concerns across organizations affects efficacy of SDP. By making use of a Federated Learning (FL) framework in combination with Differential Privacy (DP) for collective learning without exchanging any raw data, we provide a unique solution to this problem. The study makes use of advanced deep learning architectures the baseline Neural Network (NN), the sequential Long Short-Term Memory (LSTM) and the contextual Bidirectional LSTM (BLSTM) that were trained on the ten different, non-IID client datasets (software projects). Data processing included logarithmic transformation, Min Max Scaling, while the Synthetic Minority Over-sampling Technique (SMOTE) algorithm was used to address local class imbalance. The application of DP noise ensured the privacy of the model updates by quantifying them before aggregation using FedAvg. The results are a definite success for recurrent models with the BLSTM network coming on top, achieving a peak Global Accuracy of 99.05% and an F1-Score of 98.9% with the constraint of FL-DP. In comparison, the centralized, non-private BLSTM had slightly better performance in terms of accuracy, 99.3%, proving that the FL-DP framework is able to preserve close to optimal performance while increasing data security drastically. These results confirm that FL-DP is a powerful, safe and open-source paradigm for collaborative SDP, which enables a base solution for organizations that need to achieve high predictive power while maintaining high data confidentiality.

Extending ABAC Authorisation for Complex Applications: Negative Permissions, Hierarchical Attribute Values, Break Glass Override

2026-06-16T11:15:23+08:00

Much of the commercial and research writings on healthcare Real-Time Location Systems (RTLS) have concentrated on the considerable benefits of this technology for locating assets, patients and staff, with many applications developed, particularly for asset tracking. Less has been written about who is authorised to query and analyse healthcare RTLS data, including historical location data. We categorise use cases for the interactions which have been proposed and indicate further desirable authorisations which could be provided by extending Attribute Based Access Control (ABAC). The ABAC extensions we propose are: to authorise access to information protected at increasing levels of security by invoking break-glass overrides; to enable the denial of access through the use of separately-specified negative permissions, possibly at different levels of security; and supporting hierarchically-structured attribute values for users (e.g., teams and roles), operations and protected objects. We describe algorithms for rule selection and processing, largely at the conceptual level. They take account of architect-specified attribute weightings, the order of rule evaluation, and query modification if the database forms the basis for implementation. We show that our sorting strategy can reduce rule search space by over 90% compared to a naïve policy evaluation, while enabling fine-grained emergency access that current ABAC frameworks lack. We illustrate with an example of technician teams locating and maintaining sophisticated medical equipment in a large healthcare setting. Privacy, confidentiality and security become especially sensitive if technicians need to inspect medical data produced by the equipment, maybe in emergency situations. When fully implemented, our approach does not require additional application programming to provide the extended authorisation functionality described.