Speckle Noise Reduction and Image Enhancement of Echocardiographic Images for Early Detection of Heart Diseases

Abstract :

The use of ultrasound imaging in medical diagnosis is well established due to its non-invasiveness, low cost, capability of forming real time imaging, and continuing improved image quality. However, in taking ultrasound images there is still a noise that causes the decreases of the ultrasound image quality, one of the most disturbing one is speckle noise. Speckle noise has long been known as a factor that limits the detection of ultrasound images, especially in low contrast images. There are several methods that can be used in reducing noise produced by ultrasound images, in this paper the method used is Hybrid Oriented Speckle noise Reducing Anisotropic Diffusion (HOSRAD) which is a combination of the Relaxed Median Filter method and the Oriented Speckle Reducing Anisotropic Diffusion (OSRAD) method. And if the doctor wants to obtain the edges of heart boundaries and heart valves movement, the method used is Canny Edge Detection. There are several parameters used as quantitative statistical analysis in order to see the differences in ultrasound images that has been restored with the original ultrasound images used as a reference, namely Signal to Noise Ratio (SNR), Peak Signal to Noise Ratio (PSNR), and Mean Square Error (MSE), Figure of Merit (FOM), and Structural Similarity quality Index (SSIM). After various experiments, the results for the best image restoration with a gamma value of 0.5 at the gamma correction process, a time step of 0.1 at the OSRAD process, and upper and lower boundaries of 5 and 9 at the Relaxed Median Filter. So, we get the results of the HOSRAD method with MSE and PSNR values of 70.13 and 29.67 with 5 iterations. The results of the proposed method are proven to be able to improve the quality of heart images which from the beginning had MSE and PSNR of 76,123 and 29.31.

Introduction

My final project (thesis) to completed my study. The project is to reduce speckle noise found in echocardiographic images. The purposes of the project is for helping doctor/clinician to increasing the quality of the images for better view images of the heart. So, it reduces the risk of the doctor to false interpretation the images.

In this project, i am using 3 methods of image processing :

• Power Law Transform
• Oriented Speckle Reducing Anisotropic Diffusion (OSRAD)
• Relaxed Median Filter

For OSRAD and Relaxed Median Filter are combined to be Hybrid Oriented Speckle Reducing Anisotropic Diffusion (HOSRAD).

I made python file for each methods to be used as a library. I'm not yet build a GUI for this project.