Adaptive Multiscale Wavelet Compression for Preserving Critical Diagnostic Features in Medical Imaging

X-ray., MRI and CT scans are indispensable tools for diagnosing and treatment planning but the growth of the number of HR images is a problem in terms of storing and transmitting them. These challenges, therefore, require efficient compression techniques that will help overcome them, but at the same time maintaining key diagnostic details. In this paper, we introduce an adaptive multiscale wavelet-based method for compression of medical images. Our method uses a novel ROI detection algorithm along with a content-adaptive compression approach that compresses the different ROIs by a different level of compression since they have different diagnostic significance. Our approach was able to achieve a compression ratio of twenty to one while only breaking down 98. The quantitative analysis yielded insignificantly different structural similarity index (SSIM) of 5% for the ROI and 95 % for non-ROI area. There were no statistically significant difference between affirmative and negative images as identified by radiologists F1 scores for compressing images were statistically equal to F1 scores of original images p> 0. 05. It minimized on storage by 85% and transmission times by 78% than the uncompressed images. The research work suggested adaptive compression method provides a significant solution for bio signals and medical imaging data storage as planning, organizing and controlling the overwhelming increased usage of data in clinical and telemedicine environment.