Adaptive diffusion filter and absorbing light function for effective preprocessing in fetal ultrasound

The “noisy and blurry” nature of ultrasound data makes the rendering of the data a challenge in comparison with the handle of MRI and CT images. In this paper, we explore the techniques of preprocessing 3D fetal ultrasound data. An adaptive modified diffusion filter (AMDF) with local information of the image data is designed in smoothing the noisy sonogram. To remove the large noise that occludes the desired object and cannot be depressed by diffusion filtering, we propose an effective new method, called absorbing light function, to take away the occluding area interactively. These two functions are particularly useful in reducing speckle noise and removing area obstacles on 3D fetal sonogram. Thus the quality of subsequent visualization is significantly elevated