Colon Cancer Image Enhancement using Grundwald Letnikov Fractional Order Derivative

Abstract

Fractional calculus, which deals with derivatives and integrals of non-integer order, has gained significant attention over the past decade due to its ability to model complex systems more accurately than traditional integer-order calculus. Many systems in science and engineering exhibit behaviors that cannot be fully described by integer-order derivatives and integrals alone. Fractional-order methods have found significant application in image enhancement, deniosing, texture analysis, image fusion and so on. The first-order (gradient) and second-order (Laplacian) derivatives, are sensitive to abrupt changes in pixel intensities, typically corresponding to edges and corners in images. Fractional-order derivatives can suppress noise more effectively while preserving important image features, leading to smoother and more accurate results in noisy imagesIn this paper, we introduce a novel image processing method utilizing the Grundwald-Letnikov fractional order derivative and applied on EBHI-SEG dataset colon cancer images. This approach effectively suppresses noise and minimizes abrupt changes in pixel intensities, which are typically associated with edges and corners in images. The performance of the proposed method was evaluated using the PSNR (Peak Signal-to-Noise Ratio), NAE (Normalized Absolute Error), and SSIM (Structural Similarity Index) metrics. Our method demonstrated superior accuracy compared to existing techniques.