Scholarlink

<p>Biomedical image processing plays a crucial role in disease diagnosis, treatment planning, and clinical decision-making. However, the inherent challenges of noise, poor contrast, and complex structures often limit the accuracy of traditional image analysis techniques. Metaheuristic algorithms have emerged as powerful optimisation tools that mimic natural or social phenomena to solve such complex problems efficiently. This review investigates the recent advancements, advantages, and limitations of metaheuristic algorithms applied in biomedical image processing.&#xa0;A systematic literature search was conducted across Scopus, PubMed, IEEE Xplore, and SpringerLink databases for studies published between 2019 and 2025. Articles focusing on the application of metaheuristic algorithms in biomedical image segmentation, feature extraction, image registration, and disease detection were included. Data were extracted and analysed based on algorithm type, biomedical application area, performance metrics, and observed challenges.&#xa0;The review identified over 72 recent studies implementing metaheuristic algorithms such as Genetic Algorithm, Particle Swarm Optimisation, Grey Wolf Optimisation, Firefly Algorithm, Whale Optimisation, and hybrid variants. These algorithms have demonstrated superior robustness and adaptability in handling complex and noisy biomedical images. Metaheuristics significantly improved segmentation accuracy, feature selection efficiency, and disease classification performance across multiple datasets. Nonetheless, challenges remain in computational cost, parameter tuning, scalability, and convergence stability, particularly for high-dimensional or real-time biomedical data.&#xa0;Metaheuristic algorithms offer promising solutions for optimising biomedical image analysis through flexible, adaptive, and data-driven mechanisms. Despite their effectiveness, achieving optimal generalisation and reducing computational complexity remain active research areas. Future work should focus on hybrid frameworks that integrate metaheuristics with deep learning and fuzzy logic for interpretable, accurate, and scalable biomedical image processing applications.</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

Advantages and challenges of metaheuristic algorithms in biomedical image processing: a systematic review

  • 期刊论文

  • 发表日期：

    2026年5月16日

查看全文
  • Manisha Kowdiki,
  • Vaidehi Deshmukh,
  • Arti Khaparde

摘要

Biomedical image processing plays a crucial role in disease diagnosis, treatment planning, and clinical decision-making. However, the inherent challenges of noise, poor contrast, and complex structures often limit the accuracy of traditional image analysis techniques. Metaheuristic algorithms have emerged as powerful optimisation tools that mimic natural or social phenomena to solve such complex problems efficiently. This review investigates the recent advancements, advantages, and limitations of metaheuristic algorithms applied in biomedical image processing. A systematic literature search was conducted across Scopus, PubMed, IEEE Xplore, and SpringerLink databases for studies published between 2019 and 2025. Articles focusing on the application of metaheuristic algorithms in biomedical image segmentation, feature extraction, image registration, and disease detection were included. Data were extracted and analysed based on algorithm type, biomedical application area, performance metrics, and observed challenges. The review identified over 72 recent studies implementing metaheuristic algorithms such as Genetic Algorithm, Particle Swarm Optimisation, Grey Wolf Optimisation, Firefly Algorithm, Whale Optimisation, and hybrid variants. These algorithms have demonstrated superior robustness and adaptability in handling complex and noisy biomedical images. Metaheuristics significantly improved segmentation accuracy, feature selection efficiency, and disease classification performance across multiple datasets. Nonetheless, challenges remain in computational cost, parameter tuning, scalability, and convergence stability, particularly for high-dimensional or real-time biomedical data. Metaheuristic algorithms offer promising solutions for optimising biomedical image analysis through flexible, adaptive, and data-driven mechanisms. Despite their effectiveness, achieving optimal generalisation and reducing computational complexity remain active research areas. Future work should focus on hybrid frameworks that integrate metaheuristics with deep learning and fuzzy logic for interpretable, accurate, and scalable biomedical image processing applications.