Shira Nemirovsky-Rotman (Technion)
Tuesday, 19.4.2016, 12:30
Many medical images are produced daily. Compression thus plays a major role in storage and transmission applications, while enhancement, such as image de-noising, is important for diagnosis purposes. Requirements could often be conflicting, however. In order to mitigate this limitation we suggest to decompose the images into several components, where each component is distinctly described by an appropriate model and accordingly handled. For the case of ultrasound imaging, for example, we propose to describe the images as the sum of an underlying tissue pattern, speckle noise, and bright areas at the image that consist of important diagnostic information, termed 'strong reflectors'. These strong reflectors usually appear at interfaces between different tissues. They may be separated from the image and represented in a near-lossless or lossless manner. The remaining image is interpolated using a suitable model in order to improve its representation and compressibility. Speckle reduction may be applied in order to improve the image quality. Our results show that using this approach, it is possible to obtain compression ratios comparable to existing encoding methods whilst improving the image SNR and its diagnostic value.