Printer Friendly

HYBRID COMPRESSION OF MEDICAL IMAGES BASED ON LAPPED BIORTHOGONAL TRANSFORM and DISCRETE COSINE TRANSFORM.

Byline: A. Younus, G. Raja and A. K. Khan

ABSTRACT

This paper describes a new hybrid image compression technique based on Lapped Biorthogonal Transform (LBT) and Discrete Cosine Transform (DCT) for medical images. The implementation consists of image partitioning module, average gray level estimator, Region of Interest (ROI) extractor, gray level comparator and transformation module. The medical image is partitioned into 8x8 blocks and gray level estimator calculates average gray level for each block of an input image. A threshold has been devised to partition the image into ROI and non-ROI parts by using gray level comparator. Further, we have used LBT to ROI part and DCT to non-ROI part, in implementing our proposed hybrid technique. Results show that better Peak Signal to Noise Ratio (PSNR) with acceptable Compression Ratio (CR) has been achieved using hybrid scheme based on DCT-LBT as compared to the DCT-Wavelet hybrid scheme and conventional DCT or LBT individually.

Keywords: Medical image processing, Hybrid compression, Region of interest, Discrete cosine transform, Lapped biorthogonal transform.

INTRODUCTION

In recent years, the emergence of Picture Archiving and Communication System (PACS) and Hospital Information System (HIS) has brought many benefits in the field of telemedicine (Min and Sadleir,2010). As a result, doctors can share medical images across multiple locations and diagnose a patient more precisely. In addition, patients can also have option to get specialist medications from anywhere in the world. These cost effective opportunities are dependent on reliable, error-free and speedy delivery of medical images among different hospitals at various locations and medical image compression is thus required to obtain aforesaid benefits.

The images can be compressed either by use of lossy or lossless compression methods. Lossy compression techniques can achieve higher Compression Ratio (CR) at the cost of Peak Signal to Noise Ratio (PSNR) (Cardoso and Saniie, 2004). Moreover, compression error by lossy schemes adds diagnostic difficulties for a medical doctor. Therefore, lossless compression is widely used for compression of medical images as it allows exact reconstruction of images without any loss of information. However, accurate reconstruction of medical images is achieved at the cost of low CR (Oh et al., 2007). Recently, hybrid schemes are being used for medical images to achieve benefits of both lossy and lossless compression techniques.

In hybrid compression schemes, medical image is partitioned into diagnostic and non-diagnostic regions. The diagnostic part is termed as Region of Interest (ROI) and non-diagnostic part as non-ROI. Lossless and lossy compression technique is applied on ROI and non-ROI parts respectively (Kaur et al., 2011). This results in accurate reconstruction of ROI without any loss of information. Consequently, the overall compressed medical image can have higher PSNR in comparison to lossy methods and better CR in comparison to lossless techniques.

A new hybrid compression technique for medical images by use of Lapped Biorthogonal Transform (LBT) (Malvar, 1998) and Discrete Cosine Transform (DCT) (Mohammed and Abd-Elhafiez, 2009) in ROI and non-ROI respectively is applied in the proposed research. The rest of the paper is organized as follows: Section 2 describes working of LBT and DCT transforms while implementation of the proposed hybrid method is elaborated in Section 3. The simulation results are discussed in Section 4 and paper is concluded in Section 5.

An overview of DCT and lbt transforms: DCT and LBT are employed as transformation modules in JPEG (Oh and Besar, 2003) and recently introduced JPEG-XR (Maalouf and Larabi, 2009) image compression standards respectively. The brief description of DCT and LBT is as follows:

A.Discrete Cosine Transform: The DCT is widely used for image compression and it is an excellent processing tool to perform image analysis. DCT basically works by transforming an image into different frequency levels and less important frequencies are neglected during compression process to achieve higher CR (Ahmed et al., 1974). As a result compression standard using DCT is termed as lossy (Gonzalez et al., 2009). During encoding

REFERENCES

Ahmed, N., T. Natarajan, and K. R. Rao. Discrete cosine transform. IEEE Trans. on Comp. Systems, C-23:90-93 (1974).

Cardoso, G., and J. Saniie. Performance evaluation of DWT, DCT, and WHT for compression of ultrasonic signals. IEEE International Ultrasonics, Ferroelectrics and Frequency Control, 2314-2317 (2004).

DICOM, Medical Image Sets., available online at http://pubimage.hcuge.ch:8080, Jan, 2011.

Frederic, D., G. Sullivan and E. Touradj. The JPEG XR image coding standard. IEEE Signal Process. Magazine, 26:195-199 (2009).

Gonzalez, R. C., R. E. Woods, and S. L. Eddins. Digital Image Processing Using Matlab. 2nd Ed. USA (2009). image compression based on region of interest. IJCST International Conference on Image Compression; 117-118 (2011).

Maalouf, A., and M. C. Larabi. Low-complexity enhanced lapped transform for image coding in JPEG XR / HD photo. 16th IEEE International Conference on Image Processing; 5-8 (2009).

Malvar, H. S., Biorthogonal and nonuniform lapped transforms for transform coding with reduced blocking and ringing artifacts. IEEE Trans. Image Processing, 46:1043-1053 (1998).

Min, Q., and R. J. T. Sadleir. A hybrid lossless compression scheme for efficient delivery of medical image data over the internet. IEEE Second International Conference on Computer Modeling and Simulation; 319-323 (2010).

Mohammed, U. S, and W. M. Abd-Elhafiez. New approaches for DCT-Based image compression using region of interest scheme. Intl J. App. Math and Inform Sciences, 2:201-206 (2009).

Oh, T. H., and R. Besar. JPEG2000 and JPEG: image quality measures of compressed medical images. IEEE 4th National Conference on Telecommunication Technology, Shah Alam, Malaysia, 31-35 (2003).

Oh, T. H., H. S. Lim, and S. Y. Pang. Medical image processing: from lossless to lossy compression. Proceedings of IEEE TENCON Region Conference, 1-6 (2007).
COPYRIGHT 2013 Asianet-Pakistan
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Publication:Pakistan Journal of Science
Date:Jun 30, 2013
Words:927
Previous Article:OPTIMIZATION OF GERMINATION EFFECT ON FUNCTIONAL PROPERTIES OF MUNGBEAN FLOUR BY RESPONSE SURFACE METHODOLOGY.
Next Article:STONES OF BRITISH COLONIAL LAHORE: A STUDY OF TWO REMARKABLE BUILDINGS OF EARLIER PERIOD OF BRITISH COLONIAL ARCHITECTURE.
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters