Sailfish-Inspired Optimization for Pancreatic Tumor Segmentation: Introducing OptiSeg-SFO

Abstract

With advancements in radiology and computer technology, medical imaging diagnosis is transitioning towards precision and automation. Pancreas segmentation, due to the complex anatomy surrounding the pancreatic tissue and the requirement for extensive clinical expertise, stands to benefit significantly from assisted segmentation systems, enhancing clinical efficiency. However, existing segmentation models often struggle with poor generalization across images from different hospitals.This paper presents an end-to-end data-adaptive pancreas segmentation system designed to overcome the limitations posed by insufficient annotations and the lack of model generalizability.OptiSeg-SFO is a new image segmentation algorithm inspired by sailfish hunting strategies. It combines Sailfish-inspired Optimization (SFO) with adaptive methods to find the best threshold values for accurate segmentation.OptiSeg-SFO shows promise for precise and efficient image segmentation, particularly in medical imaging tasks like pancreatic CT imaging.