Medical Image Segmentation Using Combined Level Set and Saliency Analysis

Abstract

In the realm of computer vision, image segmentation has become a crucial task with widespread applications, particularly in medical imaging. Although there have been significant advancements in image segmentation methods, challenges persist in accurately delineating intricate structures within noisy and varied medical images. In this study, we have developed a novel segmentation model that combines the distance regularized level set evolution (DRLSE) model with a local gradient flow-based image (LGFI) and saliency maps. This innovative fusion addresses the limitations of existing methods and offers robust and precise solutions for medical image segmentation. We provide comprehensive mathematical formulations and demonstrate the effectiveness of the proposed model across diverse medical images. Through quantitative and qualitative analyses of the brain tumor segmentation (BraTS) 2019 dataset, we have demonstrated the superior accuracy, robustness, and computational efficiency of the proposed model in comparison with the state-of-the-art methods. This research marks a significant step toward enhancing medical image analysis, with potential applications in diagnostics and healthcare practices.

Keywords

• Active contour
• biomedical image processing
• distance regularized
• gradient-flow
• image segmentation
• saliency