P Anandan et al., International Journal of Emerging Trends in Engineering Research, 8(7), July 2020, 3760 - 3765 3760 ABSTRACT Now- a- days, medical field plays a very crucial role in our daily life, as a part of it MRI (Magnetic resonance imaging) scans, CT (computed tomography) images, Ultrasound images etc. of the victim which are one of the main things that are to be determined correctly based on which the patient’s condition is concluded and treated. The main problem here occurs is for the original image where the image gets noisy and the features of the original image are lost due to many factors. So, here in our paper, we instigate the method of image denoising technique which helps to eliminate the noisy observations and other disturbances and reconstructs the original image very accurately. The image denoising is one of the important preprocessing steps in medical field image processing analysis. For this denoising method, we are going to use the Fast Discrete Curvelet Transform which is a multi-scale geometric transform and is designed to signify the image or video sequences at different scales and angles. also the performances of it by using fast Fourier discrete curve-let transform which is based on ridge-let analysis theory for denoising procedures and makes recommendations with the help of adaptive threshold algorithm which is applied on the image and gets the original image with effectiveness also retrieves the important detail features in the image and also the quality of the image to be recovered by using the parameters like Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE). Key words: Curvelet transform, Adaptive Thresholding, PSNR, MSE, Ridgelet analysis, denoising, MRI, CT, Ultrasound image. 1. INTRODUCTION In image processing techniques, one of the fundamental challenges is image denoising in which the main aim is to get the original Image by suppressing all the noise from the noise-contaminated version of the image. This image denoising technique is mainly used and focuses on the area of medical applications and it plays an important role in an extensive range of applications such as image registration, image classification, and image segmentation, in which we obtain the unique image content, is decisive for improved performance or operation[15]. This noise reduction problem for images is to eliminate the noise from an image to recover a clean and clear image .in this the image acts as the input which is the sum of relevant pixels of the unclean image(noisy image) corresponding to the pixels of the clean image(noise-free).For example, applied denoising methods to deblurring, inpainting, and demosaicing similarly many cases can be obtained from with strength of noise and distribution of it over the image. The less computational complexity algorithm proposed by Qiang Guo uses nonlocal self-similarity and low-rank approximation (LRA) which includes a grouping of similar image patches to be low-rank. Then factorization of each group is done by using Singular Value Decomposition(SVD) [1][9].Back projection is performed to cross-verify grouping errors if any and ultimately de-noised image is formed by aggregating the factorized groups.LRA in SVD makes the algorithm less computational complex by avoiding representation of image patches. Due to the optimal energy compaction property of SVD,this algorithm provides quantitative quality results in peak signal to noise ratio and feature-similarity index parameters. Taking the advantage of nonlocal redundancy and LRA by using a block-matching technique to construct matrices of the low-rank group for weakening the noise makes the algorithm simple and efficient technique for image de-noising. (An efficient SVD based method for image denoising). Medical image denoising using convolutional denoising Auto encoders.In which the medical image denoising, though deep learning techniques are limited to sample size and computations [2][12].Gondara done an experiment with small sample size constructed using convolutional layers for competent medical image denoising. The denoising performance is increased by boosting the sample size by Medical Image Denoising using Fast Discrete Curvelet Transform P Anandan 1 , A Giridhar 2 , E Iswarya Lakshmi 3 ,P Nishitha 4 1 Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, Tamilnadu, India, anand.phd.dip@gmail.com 2 Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, Tamilnadu, India, 3 Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, Tamilnadu, India, 4 Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, Tamilnadu, India, ISSN 2347 - 3983 Volume 8. No. 7, July 2020 International Journal of Emerging Trends in Engineering Research Available Online at http://www.warse.org/IJETER/static/pdf/file/ijeter139872020.pdf https://doi.org/10.30534/ijeter/2020/139872020