Lossy image compression for digital medical imaging systems

Paul Wilhelm, David R. Haynor, Yongmin Kim, Alan C. Nelson, Eve A. Riskin

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

This paper describes a protocol for subjective and objective comparison of compressed/decompressed images to the originals and presents the results of its application to four representative and promising compression methods. Vector quantizxation is theoretically capable of producing the best compressed images, but has proven difficult to implement. Fractal compression is a relatively new compression technique, but has produced satisfactory results while beuing computationally simple. Discrete cosine transform techniques reproduce images well, but have traditionally been hampered by the need for intensive computing to compress and decompress images. To evaluate the results of each of these methods, radiologist's evaluations of image fidelity were compared to calculations of mean ssuare error, normalized mean square error, percentage mean square error, and fractal normalized mean square error for each compression method and bit rate.

Original languageEnglish
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
EditorsYongmin Kim
Pages348-358
Number of pages11
StatePublished - 1990
EventMedical Imaging IV: Image Capture and Display - Newport Beach, CA, USA
Duration: 4 Feb 19905 Feb 1990

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume1232
ISSN (Print)0277-786X

Conference

ConferenceMedical Imaging IV: Image Capture and Display
CityNewport Beach, CA, USA
Period4/02/905/02/90

Fingerprint

Dive into the research topics of 'Lossy image compression for digital medical imaging systems'. Together they form a unique fingerprint.

Cite this