Regresar

Evaluation of noise reduction filters in medical image processing usijrig OpenMP

Abstract:

Most of the medical diagnostic requires studies of medical images for to give an accurate treatment. Therefore is important the improvement of the medical images in terms of noise, quality, and morphological definition. The medical images series contains approximately 20% of noise caused by the equipment itself, especially in the x-ray modality. Therefore, the principal aim of this project is to develop an algorithm that helps suppress the noise as a preprocessing stage before medical analysis. The algorithm for noise reduction proposed uses classic and optimized mean filter, Gaussian filter, and median filter. This project also uses OpenMP parallel programming to optimize processing time and computational resources. The parallel implementation results of algorithms with sequential and classic implementation show great performance in the quality of the time processing, noise localization, and noise reduction. This improvement helps medical professionals get better details about the different pathologies for effective diagnostics and treatment.