Utilizing Cosine Wavelet Geometric Guided Sparse Representation Transform for Lossless Medical Image Compression to Enhance Image Quality
P. Kirubanantham, Ajanthaa Lakshmanan, U. V. Anbazhagu, B. Prakash, N. Ahana Priyanka · 2025
Massive-scale implementations of telemedicine and medical imaging are increasingly common today. As a result of the proliferation of medical image transmission and storage applications, adequate transmission bandwidth and memory capacity have become scarce resources. To mitigate these concerns, compression techniques are applied. Lossless image compression aims to improve accuracy, reduce bit rate, and enhance compression efficiency to enable the transmission and storage of diagnostic medical images while maintaining an acceptable level of image integrity. In this chapter, we present a method for lossless compression of medical images by employing a geometric sparse representation transform guided by cosine waves. The proposed image compression system consists of three fundamental modules: (i) segmentation, (ii) compression, and (iii) decompression. Initially, the input medical image is processed with a Perform Max Norma filter before being segmented into regions of interest (ROI) and non-ROI using a seeded region growing watershed algorithm. Subsequently, the cosine wavelet geometric guided sparse representation transform is used to compress both the ROI and non-ROI components. The compressed image is obtained by combining the compressed ROI and non-ROI. During the decompression phase, the original medical image is reconstructed by applying the reverse procedure. Experiments conducted using a variety of medical images demonstrated that the proposed approach outperforms alternative methods.