TY  - JOUR
AU  - Jalali-Varnamkhasti, Mehdi
AU  - Varnamkhasti, Mohammad Jalali
PY  - 2026
DA  - 2026/09/03
TI  - Firefly Algorithm for Medical Image Segmentation: A Systematic Review of Methods, Applications, and Future Directions
JO  - ICCK Journal of Image Analysis and Processing
T2  - ICCK Journal of Image Analysis and Processing
JF  - ICCK Journal of Image Analysis and Processing
VL  - 2
IS  - 3
SP  - 190
EP  - 205
DO  - 10.62762/JIAP.2026.507953
UR  - https://www.icck.org/article/abs/JIAP.2026.507953
KW  - firefly algorithm
KW  - medical image segmentation
KW  - metaheuristic optimization
KW  - multilevel thresholding
KW  - medical imaging
KW  - deep learning
AB  - The firefly algorithm (FA) is a nature-inspired metaheuristic increasingly applied to medical image segmentation because of its flexible search mechanism and optimization capability. This paper systematically reviews FA-based approaches for medical image segmentation and related image-analysis tasks published between 2010 and 2026. A structured search of Scopus, PubMed, IEEE Xplore, Web of Science, and Google Scholar was conducted using predefined eligibility criteria. The included studies were categorized into multilevel thresholding, entropy-based segmentation, clustering-assisted methods, adaptive and chaotic FA variants, and hybrids with deep learning and other metaheuristics. Major applications included brain, lung, breast, and microscopic image analysis. The reviewed studies generally reported competitive performance and demonstrated FA's flexibility in optimizing thresholds, clustering centers, feature subsets, and model hyperparameters. However, direct quantitative comparisons remain difficult because of heterogeneity in datasets, imaging modalities, evaluation metrics, and experimental protocols. Major limitations include computational overhead, parameter sensitivity, inconsistent benchmarking, limited clinical validation, and insufficient investigation of three-dimensional and multimodal imaging. Future research should emphasize self-adaptive FA, GPU-accelerated and parallel implementations, integration with neural architecture search and explainable artificial intelligence, extension to three-dimensional and multimodal imaging, and validation on standardized multicenter datasets. Overall, FA represents a promising optimization framework for medical image segmentation, although methodological standardization, computational improvements, and rigorous clinical validation are required for broader practical deployment.
SN  - 3068-6679
PB  - Institute of Central Computation and Knowledge
LA  - English
ER  - 

@article{JalaliVarnamkhasti2026Firefly,
  author = {Mehdi Jalali-Varnamkhasti and Mohammad Jalali Varnamkhasti},
  title = {Firefly Algorithm for Medical Image Segmentation: A Systematic Review of Methods, Applications, and Future Directions},
  journal = {ICCK Journal of Image Analysis and Processing},
  year = {2026},
  volume = {2},
  number = {3},
  pages = {190-205},
  doi = {10.62762/JIAP.2026.507953},
  url = {https://www.icck.org/article/abs/JIAP.2026.507953},
  abstract = {The firefly algorithm (FA) is a nature-inspired metaheuristic increasingly applied to medical image segmentation because of its flexible search mechanism and optimization capability. This paper systematically reviews FA-based approaches for medical image segmentation and related image-analysis tasks published between 2010 and 2026. A structured search of Scopus, PubMed, IEEE Xplore, Web of Science, and Google Scholar was conducted using predefined eligibility criteria. The included studies were categorized into multilevel thresholding, entropy-based segmentation, clustering-assisted methods, adaptive and chaotic FA variants, and hybrids with deep learning and other metaheuristics. Major applications included brain, lung, breast, and microscopic image analysis. The reviewed studies generally reported competitive performance and demonstrated FA's flexibility in optimizing thresholds, clustering centers, feature subsets, and model hyperparameters. However, direct quantitative comparisons remain difficult because of heterogeneity in datasets, imaging modalities, evaluation metrics, and experimental protocols. Major limitations include computational overhead, parameter sensitivity, inconsistent benchmarking, limited clinical validation, and insufficient investigation of three-dimensional and multimodal imaging. Future research should emphasize self-adaptive FA, GPU-accelerated and parallel implementations, integration with neural architecture search and explainable artificial intelligence, extension to three-dimensional and multimodal imaging, and validation on standardized multicenter datasets. Overall, FA represents a promising optimization framework for medical image segmentation, although methodological standardization, computational improvements, and rigorous clinical validation are required for broader practical deployment.},
  keywords = {firefly algorithm, medical image segmentation, metaheuristic optimization, multilevel thresholding, medical imaging, deep learning},
  issn = {3068-6679},
  publisher = {Institute of Central Computation and Knowledge}
}