Bio-inspired optimisation algorithms in medical image segmentation: a review

Medical image segmentation (MIS) is a primary task in medical image processing, with a great application prospect in medical image analysis and clinical diagnosis and treatment. However, MIS becomes a challenge due to the noisy imaging process of medical imaging devices and the complexity of medical images. Against this backdrop, the broad success of bio-inspired optimisation algorithms (BIOAs) has prompted the development of new MIS approaches leveraging BIOAs. As the first review of BIOAs for MIS applications, we present a comprehensive review of this recent literature, including genetic algorithm, particle swarm optimisation, ant colony optimisation, and artificial bee colony for blood vessel, organ, and tumour segmentation. We investigate the image modality and datasets that are used, discuss the application status of the four algorithms in MIS and address further research directions considering the advantages and disadvantages of each algorithm.