Printer Friendly

H.264/SVC FOR MEDICAL VIDEO COMPRESSION.

Byline: F. Nazir and G. Raja

ABSTRACT: Telemedicine can play a momentous role in remote disastrous areas for example flood, earthquake, and war time etc where there is a need of opinion of doctors/experts, not physically present, to cure masses and help them out to recover. This paper provides the evaluation and performance analysis of latest video compression standard, H.264/SVC (Scalable Video Coding) for efficient archival and fast transmission of medical videos. H.264/SVC is tested on two different kinds of medical videos Brain CT scan (Computerized Tomography) and Echocardiography. Coding efficiency of H.264/SVC presenting temporal, spatial and quality scalability is compared with that of H.264/AVC (Advanced Video Coding) single layer coding for medical videos.

Experimental results have shown that SVC can grant an appropriate level of scalability without introducing artifacts that can be misinterpreted in medical diagnosis. As a result, H.264/SVC can be successfully used in medical video compression providing temporal, spatial and/or quality scalability.

Key words: Telemedicine, H.264/SVC, compression, medical video.

INTRODUCTION

The telemedicine allow medical amenities to make diagnosis without the need of medical doctor to be physically present (Pedersen et al., 2009). CT scan (Computerized Tomography), ultrasound, magnetic resonance imaging (MRI), echocardiography, and so on, are tools for the aid of doctors to diagnose and treat patients with incredible precision and swiftness. Due to large size of medical videos, enormous disk storage is required. In addition to this, transmission of high resolution medical videos at a faster rate poses a problem and during the time of emergency, it is pivotal to make diagnosis as quickly as possible. Telemedicine can be used considerably in disaster-ridden areas where there is a problem of accessibility like flood and earthquake struck areas. The technician can take video and send it online to medical panel which can diagnose and results can be sent back.

Furthermore, telemedicine applications provide the path to take advice from specialists all over the world over complex medical issues. As a result, compression of medical video is required to address the above stated issues. Moreover, it is crucial that compression produces no noteworthy loss of detail and does not generate any perceptible artifacts that could lead to wrong diagnosis. The H.264/SVC is the recent state- of-the-art video compression standard (Schwarz et al., 2007). The video produced has reduced temporal (frame rate), spatial resolution and/or lower quality, which makes SVC to provide significant compression without compromising the subjective quality of video.

It also renders the receiving devices, versatility of having variable display and computational needs from high definition equipments to low power battery operated machines, proving scalability to be a useful tool in medical video compression. Very little work has been done on the evaluation of medical videos using H.264/AVC (Yu et al., 2005). Performance analysis of the SVC has shown that it can achieve high compression ratios (Wien et al., 2007). However, analysis is done on commercial (non-medical) videos. To the best of our knowledge, until now no effort has been made in applying H.264/SVC for medical video and there is a need to evaluate this standard with respect to medical videos. This is the key inspiration for this research.

H.264/SVC OVERVIEW

H.264/SVC is an extension of H.264/AVC (Advanced Video Coding) by ITU-T and ISO/IEC JTC 1 (ITU-T, 2007; Wiegand and Sullivan, 2007). H.264/SVC allows the encoding of video into high-quality global bit stream that consist of multiple subset bit streams. The decoded video by these subset bit streams is analogous to that achieved using the H.264/AVC (Advanced Video Coding) standard in fidelity (Schwarz and Wien, 2008) SVC is beneficial than simulcast, that permits concurrent transmission of single-layer bit streams with dissimilar bit rates and picture sizes, depending upon the application. Scalability is supported at the cost of increase in bit rate required for symbolizing a specific fidelity and the tradeoff among the desired level of scalability and the coding efficiency depend upon the particular application. The basic block diagram of an SVC encoder with two spatial layers is demonstrated in Fig-1.

Input medical video can be converted to different resolutions for encoding usin the spatial decimation block. Original or the spatially decimated video is presented for motion compensation, intra prediction and base layer coding using H.264/AVC compatible encoder to generate a bit stream with spatial and temporal scalability. Quality scalability can be added to this bit stream using fidelity scalable coding block. Consequently multiple bit streams varying in spatial resolution and fidelity can be created.

Each of them represents the source contents with a specific fidelity and spatial resolution that is a part of the overall SVC bit stream and is denoted as a layer. They are differentiated by a layer identifier. Subset bit stream having the least spatial resolution and fidelity is compatible with the H.264/AVC non-scalable bit stream, with the layer identifier equal to zero and is known as the base layer. The upper layers, mentioned as the enhancement layer (spatial or fidelity enhancement layer) employs the previously communicated information of the lower layer (one with a lesser identifier) for coding. Within a layer H.264/AVC design is used for single-layer coding. SVC introduces interlayer prediction methods for intra, inter and residual information, so that coding efficiency of the enhancement layer can be enhanced by taking advantage of the statistical dependencies among multiple layers.

The main feature of SVC is provision of scalability i.e. extracting parts from the global bit stream to acclimatize it to different needs of users in addition to the network circumstances. H.264/SVC offers spatial, temporal and/or fidelity (quality) scalability. Temporal scalability was already included in the H.264/AVC standard and can be accomplished by division of the access units into multiple layers (temporal base and one or more enhancement layers). Each access unit of a specific layer is encoded using only the access units of the same or a lower temporal layer for interpicture prediction. SVC introduced the features of spatial and fidelity scalability.

Spatial scalability is the case when subset bit stream supplies more than one picture size (spatial resolution) (Andrew et. al., 2007) while the fidelity scalability describes the situation when the subset bit stream delivers the source content at the original spatio-temporal resolution but with variable reconstruction quality. SVC can present diversity in combinations of the above mentioned three fundamental scalability types.

RESULTS

Medical videos were tested on JSVM 9.19.8 (Joint Scalable Video Model) (Reichel et. al., 2007) using SVC (scalable video coding). We used medical videos of Brain CT scan and Echocardiography. Both the videos are in YUV 4:2:0 format and have 4cif resolution (704x576). We encoded 200 frames of each video using the using the RATE Control OFF and QP (quantization parameter) from 25 to 40, because above 40 there is a chance of introduction of artifacts that could lead to wrong diagnosis. Parameters of analysis are PSNR (peak signal to noise ratio) for quality and bit rate. We used two layers with resolutions 4CIF (704x576) and CIF (352x288). In our experiments, only the features permitted in the Baseline Profile of H.264/AVC and Scalable Baseline Profile of H.264/SVC were considered.

Temporal Scalability: Table 1 gives the temporal scalability values at different frame rates. It shows that as quantization parameter (QP) increases, bit rate and PSNR decreases. With the increase in frame rate, increase in the bit rate is observed. Fig-2 shows the rate-distortion results for the test medical videos Brain CT scan (Computerized Tomography) and Echocardiography for the temporal scalability. All the videos are having 4CIF (704x576) resolution. Four frame rates are supported by all the videos; they are 7.5 Hz or fps (frames per second),15 fps, 30 fps and 60 fps. The results show that multiple subset bit streams varying in frame rate are supported by the global SVC bit stream from which any of the sub streams can be used for decoding.

Table-1: Temporal Scalability

Test###QP###Frame Rate (Hz)

Sequence###7.5###15###30###60

###Bit rate###PSNR###Bit rate###PSNR###Bit rate###PSNR###Bit rate###PSNR

Brain CT###25###828.9600###44.3006###857.5608###44.2089###875.4552###44.1915###886.9584###44.2691

Scan###30###578.6280###41.7112###598.4328###41.6496###615.6000###41.6318###625.2264###41.6418

###35###406.7544###38.8877###427.6416###38.8824###449.2560###38.8834###462.8544###38.8911

###40###340.8144###36.4206###370.1376###36.5426###404.6808###36.6038###434.3976###36.6385

Echocardi-###25###1048.728###43.9782###1375.4328###42.7956###1601.6328###42.2695###1671.864###42.3009

ography###30###685.6272###41.2554###881.0184###40.5062###1014.3888###40.1760###1062.9024###40.1529

###35###474.6720###39.0415###601.9608###38.4566###701.3808###38.2426###748.8504###38.2314

###40###393.2016###37.0375###520.6488###36.8060###628.0512###36.7270###686.1240###36.7397

###25###886.9584###44.2691###1197.7296###44.7919

Brain CT Scan###30###625.2264###41.6418###736.2144###41.7122

###35###462.8544###38.8911###508.6272###38.5501

###40###434.3976###36.6385###460.4656###35.3471

###25###1671.8640###42.3009###1956.3888###42.4422

Echocardiography###30###1062.9024###40.1529###1100.4072###39.8710

###35###748.8504###38.2314###787.8936###37.5688

###40###686.1240###36.7397###903.8016###35.4719

Spatial Scalability: The rate-distortion results for the medical video shown in Fig-2 depict that the PSNR achieved using H.264/SVC is high enough for correct diagnosis.

We used two spatial layers in our experiments with resolutions 4CIF (704x576) and CIF (352x288).

Table 2 enlists the experimental results for the two layer H.264/SVC coding and single layer H.264/AVC coding for 4CIF (704x576) resolution. It can be seen from the Table-2 that 2-layer spatial coding, at the same quantization parameter (QP), supplies higher bit rate as compared to single layer coding.

evident from the graphs that SVC spatial scalability is supported at the price of raise in the bit rate to some extent but in this way two bit streams are created of the two desired resolutions (4CIF and CIF) varying in coding efficiency and visual quality.

Fidelity (Quality or SNR) Scalability: Table 3 gives the results at 4CIF (704x576) resolution for both single layer and two layer coding. It depict that at the same quantization parameter (QP) single layer coding supports lesser bit rates as compared to 2-layer coding.

Table-3: Fidelity Scalability

Test Sequence###QP###Single Layer###2-Layer###

###Bit rate###PSNR###Bit rate###PSNR

###25###886.9584###44.2691###1107.6816###45.3198

Brain CT Scan###30###625.2264###41.6418###766.2288###41.8675

###35###462.8544###38.8911###609.0240###39.1703

###40###434.3976###36.6385###552.4344###36.6886

###25###1671.8640###42.3009###1984.3608###43.3116

Echocardiography###30###1062.9024###40.1529###1224.0864###40.2593

###35###748.8504###38.2314###908.4096###38.3339

###40###686.1240###36.7397###815.5752###36.7317

CONCLUSION: We have presented the performance analysis of H.264/SVC for medical videos. Coding efficiency of SVC is compared with that of H.264/AVC single layer coding for the medical videos of Brain CT scan and Echocardiography. Experimental results have revealed that SVC can supply an appropriate level of scalability without introducing perceptible artifacts that can lead to error in medical diagnosis. Even though scalable coding is achieved at some cost of bit rate, the difference between the coding efficiency of H.264/AVC single layer coding and SVC is remarkably minute. As a result, H.264/SVC can be used in medical video compression providing temporal, spatial and/or quality scalability. The future of telemedicine operatives will greatly depend upon the fact, that how quickly and efficiently the digital video are stored and transferred from one place to another.

REFERENCES

Andrew, C., Segall, and G. J. Sullivan. Spatial scalability within the H.264/AVC scalable video coding extension. IEEE Trans. on Circuits and Systems,17:1121-1135 (2007).

ITU-T and ISO/IEC JTC 1. Advanced video coding for generic audiovisual services. ITU-T Recommendation H.264 and ISO/IEC 14496-10 (MPEG-4 AVC), Version 8 (including the SVC extension): Consented in July (2007).

Pedersen, P. C., B. W. Dickson and J. Chakareski.Telemedicine applications of mobile ultrasound. IEEE Intl. Workshop on Multimedia Signal Processing, Brazil:1-6 (2009).

Reichel, J., H. Schwarz, and M. Wien, Joint Scalable Video Model 11(JSVM 11), Joint Video Team, Doc. JVT-X202 (2007).

Schwarz, H., D. Marpe and T. Wiegand. Overview of the scalable video coding extension of the H.264/AVC standard. IEEE Trans. on Circuits and Systems for Video Technology, 17:1103-1120 (2007).

Schwarz, H., and M. Wien. The scalable video coding extension of the H.264/AVC standard. IEEE Signal Processing Mag., 25:135-141 (2008).

Wiegand, T., and G. J. Sullivan. The H.264/AVC video coding standard. IEEE Signal Processing Mag.,24:148-153 (2007).

Wien, M., H. Schwarz and T. Oelbaum. Performance analysis of SVC. IEEE Trans. on Circuits and Systems for Video Technology, 17:1194-1203 (2007).

Yu H., Z. Lin and F. Pan. Applications and improvement of H.264 in medical video compression. IEEE Trans. on Circuits and Systems, 52:2707-2716 (2005).
COPYRIGHT 2011 Asianet-Pakistan
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Nazir, F.; Raja, G.
Publication:Pakistan Journal of Science
Article Type:Report
Geographic Code:9PAKI
Date:Dec 31, 2011
Words:2238
Previous Article:TEMPORAL POPULATION GROWTH OF FAISALABAD CITY.
Next Article:FLOOD FREQUENCIES IN SOAN VALLEY.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters