A Hybrid Technique for Medical Image Denoising using NN, Bilateral filter and LDA

Medical imaging technology is becoming an important component of large numbers of applications such as diagnosis, Treatment and research. Medical images like CT, X-Ray, MRI, PET and SPECT have minute information about heart, nerves and brain. These images need to be more accurate and free from distortion or noise. Noise reduction or removing plays an important role in medical images. Various methods of noise removal such as: wavelets, filters and thresholding based on wavelets. Although these types of methods produced good results but still have some drawbacks or limitation Considering and analyzing the limitations of the previous methods our research presents neural networks as an efficient and robust tool for noise reduction. In our research we use neural network as the learning algorithm which follows the supervised learning. In this paper Bilateral filter is defined for its effectiveness in edge-preserved image Denoising. Bilateral filter improves the Denoising efficiency, preserves the fine structures and also reduces the Rician noise. The LDA analyzing the limitations or drawbacks of the previous methods with proposed work. The proposed research use both mean and median statistical functions for calculating the output pixels results in terms of PSNR, MSE, And Mean SSIM. Keywords-Noising, De-noising, Medical images, NN, Bilateral filter And LDA.