Subband Thresholding for Near-Lossless Medical Image Compression

Abstract

Medical images play a prominent role in diagnostic and treatment planning in the medical arena. Medical images are acquired using several medical imaging modalities like Computed Tomography (CT), Magnetic Resonance Imaging (MRI), UltraSound (US), Positron Emission Tomography (PET), and X-Ray. These are used as an image acquisition tool to capture the medical images. Numerous volumes of medical data are produced for processing, storing and transmission over the network for telemedicine services every day. Those images are in the form of multidimensional data with high resolution. Hence, it certainly yields higher bandwidth and storage. Thus, it is very essential for medical image applications to reduce storage and to solve transmission problems. Every information it holds is a matter of human lives. The fast and safe transmission of these data will decrease the mortality rate. Lossy type of image compression technique achieves better compression rate than image quality where lossless compression can only retain the image quality. Therefore, there is an existing need for near-lossless compression which compromises both the compression rate and image quality. In this paper, we have proposed a near-lossless medical image compression technique that thresholds the sub-bands to increase the number of zero coefficients. And as an entropy encoder run-length encoding is applied to compress the medical image to retain the diagnostic information with high compression efficiency with better image quality. Compression Ratio (CR), Peak Signal to Noise Ratio (PSNR), Time Complexity and Bits Per Pixel (BPP) were used as assessment parameters to examine the performance of our proposed method. We achieved an average of 2 BPP and PSNR of 42.43 dB which is higher than the existing methods.

Image-Compression Techniques: Classical and “Region-of-Interest-Based” Approaches Presented in Recent Papers

“…The lossy compression type can achieve high compression ratios, but the image quality is affected, while the lossless compression type retains the image quality but does not have a good compression ratio. Starting from these two types of compression, the authors of paper [21] (Punitha et al) presents the idea of a near-lossless compression which can achieve satisfactory results for both compression rate and image quality. The proposed method uses sub-bands thresholds to increase the number of zero coefficients.…”