HYBRID DCT-DWT BASED ROI MEDICAL IMAGE COMPRESSION FOR TELEMEDICINE APPLICATION

Abstract
Medical imaging greatly affects medication, particularly in the fields of diagnosis analysis and careful surgical planning. In any case, medical imaging gadgets continue delivering a great deal of data information for each patient. Medical image examination and data compression is a significant region of research which targets delivering calculations that diminish record size and simultaneously keep up important symptomatic data. Medical image compression applications are quality-driven applications which request high caliber for specific areas that have demonstrative significance for diagnosis, where even little quality decrease presented by lossy coding may modify resulting finding, which may cause serious lawful outcomes. The fundamental focal point of this paper is to investigate procedures and discover a compression algorithm that can eliminate irrelevant medical data and reconstruct medical image rapidly while keeping up a decent degree of visual quality for certain regions of medication where it is adequate to keep up high image quality just for indicatively noteworthy regions, for instance, tumor segment of the cerebrum MRI. Wavelet multi-resolution decomposition of images has indicated its proficiency in many image processing areas and explicitly in compression. Because of this, The Discrete Wavelet Transform (DWT) is used to code Region of Interest and Discrete Cosine Transformation (DCT) is utilized to code background region. A couple of examinations were led to break down the calculations dependent on compression proportion, decompressed image quality and execution speed.

Authors
Nayankumar Patel1, Ved Vyas Dwivedi2, Ashish Kothari3
C U Shah University, India1, Gokul Global University, India2, Atmiya University, India3

Keywords
Medical Imaging, ROI, DWT, DCT, Lossy and Lossless Compression
Published By :
ICTACT
Published In :
ICTACT Journal on Image and Video Processing
( Volume: 11 , Issue: 2 )
Date of Publication :
November 2020
DOI :

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.