Int. J. Telemedicine and Clinical Practices, Vol. 1, No. 1, 2015 17
Copyright © 2015 Inderscience Enterprises Ltd.
Compression of medical images for remote diagnosis
based on geometric transforms
Sujitha Juliet*
Department of Information Technology,
School of Computer Science and Technology,
Karunya University,
Coimbatore-641114, India
Email: sujitha_juliet@yahoo.com
*Corresponding author
Elijah Blessing Rajsingh
School of Computer Science and Technology,
Karunya University,
Coimbatore-641114, India
Email: elijahblessing@gmail.com,
Kirubakaran Ezra
Department of Outsourcing,
Bharat Heavy Electricals Limited,
Trichy, India
Email: e_kiru@yahoo.com
Abstract: With the tremendous growth in imaging applications and the
development of filmless radiology, the need for medical image compression
becomes essential. This paper proposes a simple method for medical image
compression using geometric spatial transforms. The proposed method makes
use of the advantage of image scaling that stretches the image region by an
adjustable scaling factor to view the particular region of an image. This method
also considers the relationship between neighbouring pixels through bilinear
interpolation to preserve the visual quality of the image. The dependencies
between the geometric transformed coefficients are exploited using set
partitioning in hierarchical trees (SPIHT) encoder. Experimental results on a set
of medical images demonstrate that besides having better visual quality for the
selected region of interest, the proposed method provides competing
performance compared with the conventional and state-of-the-art image
compression methods, in terms of peak signal to noise ratio and computational
time.
Keywords: medical image compression; telemedicine; geometric transforms;
bilinear interpolation; image scaling; SPIHT.
Reference to this paper should be made as follows: Juliet, S., Rajsingh, E.B.
and Ezra, K. (2015) ‘Compression of medical images for remote diagnosis
based on geometric transforms’, Int. J. Telemedicine and Clinical Practices,
Vol. 1, No. 1, pp.17–31.