A robust Segmentation Model-based Fuzzy Swarm Intelligence and Logistic Chaotic Map for Hepatic CT Focal Lesion Segmentation

Computed tomography (CT) scans of the liver and hepatic lesions must be segmented in order to diagnose liver abnormalities precisely and reduce the likelihood of liver surgery. In this study, a dynamic hybrid model is proposed for the automatic identification and segmentation of hepatic lesions from CT scans. More specifically, a powerful optimization model for accuracy, speed, and optimal convergence based on kernel fuzzy c-means (FCM), chaotic map, and antlion optimization (ALO) algorithm for automatic hepatic focal lesion segmentation is proposed; named (CALO-FCM). In order to achieve the best cluster centroids and produce more accurate segmentation results, ALO is employed to guide FCM. By balancing exploration and exploitation rates, the performance of ALO is improved in terms of local minima avoidance and convergence speed. It might be argued that the CALO improves computational performance and prevents the FCM from becoming trapped in local minima. The proposed model shows good detection and segmentation outcomes on a set of patients with abdominal liver CT when compared to other methods. Additionally, the experimental findings demonstrated that the proposed model could locate the ideal centroids and avoid the local optima issue. This new hybrid model may lead to an earlier and more accurate clinical diagnosis of a hepatic lesion, assisting medical professionals in their judgement and enabling patients to receive an earlier prognosis.