Scholarlink

<p>Malaria remains a global health challenge, requiring innovative diagnostic approaches to improve detection and classification accuracy. This research proposes a novel optimization-based method, the levy flight-enhanced Slime Mould Algorithm (LfSMA), addressing the limitations of the original Slime mould Algorithm (SMA), such as premature convergence, local optima entrapment, and imbalance between exploration and exploitation. LfSMA is integrated with adaptive Fuzzy C-means (AFCM) and Support Vector Machine (SVM) techniques to enhance the diagnostic precision, forming the AFCM-LfSMA-SVM model. This hybrid framework combines the strengths of SMA’s simplicity and minimal parameter requirements with the robustness of Levy flight and advanced classification techniques to optimize malaria parasite detection and classification into positive and negative categories. The LfSMA integrates the levy flight to introduce significant steps, improving the algorithm’s search capabilities, escapes from local optima, and convergence efficiency. The study tested LfSMA on the CEC2021 benchmark functions. The outcome is compared with seven state-of-the-art algorithms and the results showed better performance than other compared metaheuristic algorithms. The AFCM-LfSMA-SVM model is applied to a malaria dataset obtained from health facilities in Sierra Leone, and the results demonstrate its superiority over other methods, achieving higher accuracy, specificity, Matthew’s correlation coefficient (MCC), and sensitivity. These findings indicate that the proposed approach is a reliable and effective tool for malaria diagnosis, clinical applications, and improved management.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A hybrid slime mould algorithm with Levy Flight based mutation for malaria parasite detection

  • 期刊论文

  • 发表日期：

    2025年4月16日

查看全文
  • Ibrahim Musa Conteh,
  • Aminu Onimisi Abdulsalami,
  • Gibril Njai,
  • Qingguo Du

摘要

Malaria remains a global health challenge, requiring innovative diagnostic approaches to improve detection and classification accuracy. This research proposes a novel optimization-based method, the levy flight-enhanced Slime Mould Algorithm (LfSMA), addressing the limitations of the original Slime mould Algorithm (SMA), such as premature convergence, local optima entrapment, and imbalance between exploration and exploitation. LfSMA is integrated with adaptive Fuzzy C-means (AFCM) and Support Vector Machine (SVM) techniques to enhance the diagnostic precision, forming the AFCM-LfSMA-SVM model. This hybrid framework combines the strengths of SMA’s simplicity and minimal parameter requirements with the robustness of Levy flight and advanced classification techniques to optimize malaria parasite detection and classification into positive and negative categories. The LfSMA integrates the levy flight to introduce significant steps, improving the algorithm’s search capabilities, escapes from local optima, and convergence efficiency. The study tested LfSMA on the CEC2021 benchmark functions. The outcome is compared with seven state-of-the-art algorithms and the results showed better performance than other compared metaheuristic algorithms. The AFCM-LfSMA-SVM model is applied to a malaria dataset obtained from health facilities in Sierra Leone, and the results demonstrate its superiority over other methods, achieving higher accuracy, specificity, Matthew’s correlation coefficient (MCC), and sensitivity. These findings indicate that the proposed approach is a reliable and effective tool for malaria diagnosis, clinical applications, and improved management.