| Abstrakt: |
To mitigate the noise effects without information loss at the edges of the radiological images, a well-designed preprocessing algorithm is required to assist the radiologists. This paper proposes a hybrid adaptive preprocessing algorithm that utilizes a Rudin_Osher_Fatemi (R_O_F) model for edge detection, Richardson_Lucy (R_L) algorithm for image enhancement, and block matching 3D Collaborative filtering for denoising images. The performance of the proposed method is assessed and estimated on two realistic datasets, one on chest X-ray images and the other on MRI and CT images. The proposed hybrid system verifies the data reliability of Gaussian noise-affected medical images. The simulation results show that the proposed adaptive method attains a high-value peak signal-to-noise ratio of 47.4433 dB for chest X-ray and 46.8674 dB for MRI and CT datasets, respectively, at a standard deviation value of 2. The performance analysis of the proposed scheme is further carried out using various statistical parameters, such as root-mean-square error, contrast-to-noise ratio, Bhattacharya coefficient, and edge preservation index. A comparative analysis of denoised image quality shows that the proposed system achieves better performance than several existing denoising methods. |
Full text from SpringerLink