Semi-adaptive vector quantization and its application in medical image compression

Jian Hong Hu, Yao Wang, Patrick Cahill

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, we introduce a semi-adaptive vector quantization (SAVQ) method, which is a combination of the traditional VQ scheme using a fixed code book and the locally adaptive VQ (LAVQ) method which dynamically constructs a code book according to the input data stream. The code book in SAVQ consists of two parts: a fixed part that is designed based on certain training signals as in VQ, and an adaptive part that is updated based on the input vectors to be compressed. The proposed method is more effective than VQ and LAVQ for semi-stationary signals that have patterns common over different images as well as features specific to a particular image. Such is the case with medical images, which have similar tissue characteristics over different images, as well as local variations that are patient and pathology dependent. The SAVQ as well as VQ and LAVQ methods have been applied to multispectral magnetic resonance brain images. The SAVQ has achieved higher compression ratios than the VQ and LAVQ methods over a wide range of reproduction quality, with more significant improvement in the mid to high quality range. Furthermore, under the same quality criterion, SAVQ requires a much smaller code book than VQ, making the former less time and memory demanding. Readings by neuroradiologists have suggested that images produced by SAVQ at compression ratios up to 40 (for Mlii data with 3 or 4 images/set, 256x 256 pixels/image, and 16 bits/pixel) are acceptable for primary reading.

Original languageEnglish (US)
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
Pages902-913
Number of pages12
Volume2094
DOIs
StatePublished - 1993
EventVisual Communications and Image Processing 1993 - Cambridge, MA, United States
Duration: Nov 7 1993Nov 7 1993

Other

OtherVisual Communications and Image Processing 1993
CountryUnited States
CityCambridge, MA
Period11/7/9311/7/93

Fingerprint

vector quantization
Vector Quantization
Vector quantization
Image Compression
Medical Image
Image compression
compression ratio
Pixels
pixels
Pathology
Magnetic resonance
Compression
Brain
Pixel
pathology
Tissue
brain
magnetic resonance
Data storage equipment
Magnetic Resonance

ASJC Scopus subject areas

  • Applied Mathematics
  • Computer Science Applications
  • Electrical and Electronic Engineering
  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics

Cite this

Hu, J. H., Wang, Y., & Cahill, P. (1993). Semi-adaptive vector quantization and its application in medical image compression. In Proceedings of SPIE - The International Society for Optical Engineering (Vol. 2094, pp. 902-913) https://doi.org/10.1117/12.158007

Semi-adaptive vector quantization and its application in medical image compression. / Hu, Jian Hong; Wang, Yao; Cahill, Patrick.

Proceedings of SPIE - The International Society for Optical Engineering. Vol. 2094 1993. p. 902-913.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Hu, JH, Wang, Y & Cahill, P 1993, Semi-adaptive vector quantization and its application in medical image compression. in Proceedings of SPIE - The International Society for Optical Engineering. vol. 2094, pp. 902-913, Visual Communications and Image Processing 1993, Cambridge, MA, United States, 11/7/93. https://doi.org/10.1117/12.158007
Hu JH, Wang Y, Cahill P. Semi-adaptive vector quantization and its application in medical image compression. In Proceedings of SPIE - The International Society for Optical Engineering. Vol. 2094. 1993. p. 902-913 https://doi.org/10.1117/12.158007
Hu, Jian Hong ; Wang, Yao ; Cahill, Patrick. / Semi-adaptive vector quantization and its application in medical image compression. Proceedings of SPIE - The International Society for Optical Engineering. Vol. 2094 1993. pp. 902-913
@inproceedings{85806b0919844985800164c5e88eeaf2,
title = "Semi-adaptive vector quantization and its application in medical image compression",
abstract = "In this paper, we introduce a semi-adaptive vector quantization (SAVQ) method, which is a combination of the traditional VQ scheme using a fixed code book and the locally adaptive VQ (LAVQ) method which dynamically constructs a code book according to the input data stream. The code book in SAVQ consists of two parts: a fixed part that is designed based on certain training signals as in VQ, and an adaptive part that is updated based on the input vectors to be compressed. The proposed method is more effective than VQ and LAVQ for semi-stationary signals that have patterns common over different images as well as features specific to a particular image. Such is the case with medical images, which have similar tissue characteristics over different images, as well as local variations that are patient and pathology dependent. The SAVQ as well as VQ and LAVQ methods have been applied to multispectral magnetic resonance brain images. The SAVQ has achieved higher compression ratios than the VQ and LAVQ methods over a wide range of reproduction quality, with more significant improvement in the mid to high quality range. Furthermore, under the same quality criterion, SAVQ requires a much smaller code book than VQ, making the former less time and memory demanding. Readings by neuroradiologists have suggested that images produced by SAVQ at compression ratios up to 40 (for Mlii data with 3 or 4 images/set, 256x 256 pixels/image, and 16 bits/pixel) are acceptable for primary reading.",
author = "Hu, {Jian Hong} and Yao Wang and Patrick Cahill",
year = "1993",
doi = "10.1117/12.158007",
language = "English (US)",
volume = "2094",
pages = "902--913",
booktitle = "Proceedings of SPIE - The International Society for Optical Engineering",

}

TY - GEN

T1 - Semi-adaptive vector quantization and its application in medical image compression

AU - Hu, Jian Hong

AU - Wang, Yao

AU - Cahill, Patrick

PY - 1993

Y1 - 1993

N2 - In this paper, we introduce a semi-adaptive vector quantization (SAVQ) method, which is a combination of the traditional VQ scheme using a fixed code book and the locally adaptive VQ (LAVQ) method which dynamically constructs a code book according to the input data stream. The code book in SAVQ consists of two parts: a fixed part that is designed based on certain training signals as in VQ, and an adaptive part that is updated based on the input vectors to be compressed. The proposed method is more effective than VQ and LAVQ for semi-stationary signals that have patterns common over different images as well as features specific to a particular image. Such is the case with medical images, which have similar tissue characteristics over different images, as well as local variations that are patient and pathology dependent. The SAVQ as well as VQ and LAVQ methods have been applied to multispectral magnetic resonance brain images. The SAVQ has achieved higher compression ratios than the VQ and LAVQ methods over a wide range of reproduction quality, with more significant improvement in the mid to high quality range. Furthermore, under the same quality criterion, SAVQ requires a much smaller code book than VQ, making the former less time and memory demanding. Readings by neuroradiologists have suggested that images produced by SAVQ at compression ratios up to 40 (for Mlii data with 3 or 4 images/set, 256x 256 pixels/image, and 16 bits/pixel) are acceptable for primary reading.

AB - In this paper, we introduce a semi-adaptive vector quantization (SAVQ) method, which is a combination of the traditional VQ scheme using a fixed code book and the locally adaptive VQ (LAVQ) method which dynamically constructs a code book according to the input data stream. The code book in SAVQ consists of two parts: a fixed part that is designed based on certain training signals as in VQ, and an adaptive part that is updated based on the input vectors to be compressed. The proposed method is more effective than VQ and LAVQ for semi-stationary signals that have patterns common over different images as well as features specific to a particular image. Such is the case with medical images, which have similar tissue characteristics over different images, as well as local variations that are patient and pathology dependent. The SAVQ as well as VQ and LAVQ methods have been applied to multispectral magnetic resonance brain images. The SAVQ has achieved higher compression ratios than the VQ and LAVQ methods over a wide range of reproduction quality, with more significant improvement in the mid to high quality range. Furthermore, under the same quality criterion, SAVQ requires a much smaller code book than VQ, making the former less time and memory demanding. Readings by neuroradiologists have suggested that images produced by SAVQ at compression ratios up to 40 (for Mlii data with 3 or 4 images/set, 256x 256 pixels/image, and 16 bits/pixel) are acceptable for primary reading.

UR - http://www.scopus.com/inward/record.url?scp=84892883921&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84892883921&partnerID=8YFLogxK

U2 - 10.1117/12.158007

DO - 10.1117/12.158007

M3 - Conference contribution

AN - SCOPUS:84892883921

VL - 2094

SP - 902

EP - 913

BT - Proceedings of SPIE - The International Society for Optical Engineering

ER -