A new hybrid optimization algorithm designated as Memetic-Enhanced Glowworm Swarm Optimization with Adaptive Particle Refinement achieves statistically significant improvements in data sanitization accuracy and restoration fidelity on higher-dimensional medical records within distributed cloud environments. Developed to address persistent privacy and security limitations that currently hinder large-scale adoption of cloud-based healthcare systems, the framework focuses on securing data sanitization alongside subsequent data restoration operations. Across comparative evaluations, the computational approach proved competitive with leading baselines on lower-dimensional datasets while securing measurable performance advantages when processing complex clinical datasets.
The proposed computational framework integrates multiple optimization strategies to manage privacy preservation and accurate data recovery. At its core, the architecture couples Glowworm Swarm Optimization for multimodal exploration across continuous solution domains with Dynamic Particle Swarm Optimization to accelerate convergence in discrete optimization tasks. To refine local solutions, the framework incorporates Memetic Algorithms that enhance local utilization, alongside Simulated Annealing mechanisms specifically deployed to escape local optima during search iterations. Together, these combined algorithms coordinate to achieve peak performance across data hiding and recovery procedures in cloud settings.
Beyond its combined optimization foundation, the algorithmic model introduces adaptive neighborhood scaling alongside improved key extraction methods. These structural modifications are engineered to maintain overall data integrity during sanitization and restoration procedures without incurring excessive information loss. By balancing neighborhood parameters dynamically and refining the extraction of protective keys, the framework manages the fundamental trade-offs between masking sensitive healthcare attributes and preserving the underlying clinical utility required for legitimate healthcare information sharing and routine medical access.
The experimental evaluation systematically assessed the proposed framework across three independent healthcare benchmark datasets obtained from the University of California, Irvine machine learning repository. These evaluation testbeds specifically comprised the Heart Disease dataset, the Breast Cancer Wisconsin dataset, and the Pima Indians Diabetes dataset, representing diverse diagnostic scopes and dimensional characteristics. By testing against multiple distinct clinical domains, the study measured the framework across varying attribute structures and operational conditions relevant to real-world healthcare cloud storage environments.
Researchers evaluated the hybrid system against five established benchmark optimization techniques to gauge relative performance gains. The comparative baseline suite included Artificial Bee Colony, Firefly Algorithm, standard Glowworm Swarm Optimization, and Genetically Modified Glowworm Swarm Optimization, alongside the more recent 2022 Hunter-Prey Optimization technique. To verify the independent contribution of each structural component within the hybrid system, the researchers conducted a comprehensive ablation study, which was accompanied by paired statistical significance testing and a detailed computational complexity analysis.
In comparative benchmarking, the hybrid framework demonstrated superior performance on higher-dimensional records while matching top baseline results on lower-dimensional inputs across several operational metrics. The evaluation examined sanitization accuracy, restoration fidelity, convergence rate, and stability across comprehensive radar chart multi-metric visualizations. Furthermore, the framework retained its operational robustness when subjected to substantial sensitivity variations spanning from 10 percent to 70 percent, demonstrating consistent sanitization and restoration capabilities across wide fluctuations in data sensitivity demands.
The study authors established the framework’s clear superiority and distinct advantages in balancing sanitization accuracy with data recovery fidelity under the rigorous operational constraints of medical cloud infrastructure. The experimental validation confirmed that combining adaptive particle refinement with memetic search structures successfully addresses the delicate balance between safeguarding patient privacy and maintaining retrievable data fidelity. Through systematic empirical testing against multiple optimization baselines and clinical datasets, the resulting model confirms the viability of multi-algorithm optimization for secure, privacy-preserving healthcare cloud operations.
Subject of Research: Medical cloud data privacy and swarm optimization
Article Title: A Memetic-Enhanced Glowworm Swarm Optimization Algorithm with Adaptive Particle Refinement for secure and privacy-aware medical cloud computing
Article References: Javed, M., Himanshu, Tripathy, S., Arora, H., Kochar, T., Tyagi, H., Singh, P., & Solke, N. (2026). A Memetic-Enhanced Glowworm Swarm Optimization Algorithm with Adaptive Particle Refinement for secure and privacy-aware medical cloud computing. Journal of Cloud Computing. https://doi.org/10.1186/s13677-026-00988-1
Image Credits: AI Generated
DOI: 10.1186/s13677-026-00988-1
Keywords: cloud computing, medical data privacy, data sanitization, data restoration, glowworm swarm optimization, dynamic particle swarm optimization, memetic algorithms, simulated annealing, adaptive particle refinement, key extraction, Heart Disease dataset, Breast Cancer Wisconsin dataset, Pima Indians Diabetes dataset, Hunter-Prey Optimization, Firefly Algorithm, Artificial Bee Colony, computational complexity analysis
News Source: Denise Maddox. (November 19, 2026). Hybrid Glowworm Algorithm Improves Medical Cloud Data Sanitization and Restoration Fidelity. Scienmag.



