A TBD Fuzzy C-Means Clustering Algorithm Based on Quadratic Polynomial for Cardiac Image Segmentation

In this study, we propose a TBD fuzzy C-means clustering algorithm based on quadratic polynomial for segment cardiac images. Our approach utilizes Ruban's total bregman divergence (TBD) to capture global information effectively, reduce noise and differences in values, and exhibit higher flexibility and applicability across various models. By employing the fuzzy C-means clustering algorithm with a quadratic polynomial surface as the clustering center, our method achieves efficient segmentation of weak edges in images affected by weak noise. Furthermore, we leverage spatial information from neighboring pixels for local processing to enhance the anti-noise robustness of our algorithm. The convergence of our proposed algorithm is verified using Zangwill's theorem and boundary-based Hessian theorem. Experimental results demonstrate that our approach not only effectively segments images but also exhibits high accuracy even in the presence of certain levels of noise or non-uniformity.