Medical Image Denoising Using Artificial Intelligence–Based Algorithms

Noise reduction in medical images is one of the main challenges faced by researchers. Classical denoising methods can reduce image noise; however, they sometimes degrade image quality and lose important information, such as by blurring image edges. To address this challenge, this work proposes an optimized filter based on the Generalized Cauchy (GC) distribution and an artificial intelligence–based algorithm that preserve image details while reducing noise. The Generalized Cauchy filter and the bilateral filter are two parameter-based filters that significantly suppress image noise. Parameter-based filters require proper parameter selection to effectively remove noise while preserving edge details. For this purpose, one filter is considered. In most AI-based bilateral filters, only the range and spatial parameters typically based on the Gaussian distribution are optimized, while the neighborhood radius remains fixed. The results of the filter on various noisy images show that the neighborhood radius has a major effect on filter performance. Since the filter designed with the GC function achieves considerable noise reduction and exhibits behavior similar to the Gaussian function, it is combined with the bilateral filter to design a filter proposed in this paper. The range, spatial parameters, and neighborhood radius are optimized using PSNR as the fitness function. With the help of AI-based optimization algorithms, namely the Gaining-Sharing Knowledge based Optimization Algorithm (GSK), the parameters of the bilateral filter are optimized. Finally, the performance of the proposed filters is evaluated on medical images contaminated with Gaussian noise. The performance of these filters is compared with that of classical and other AI-based filters. Experimental findings demonstrate that the proposed filters outperform the others.