International Journal of Electrical and Computer Engineering (IJECE) Vol. No. June 2016, pp. ISSN: 2088-8708. DOI: 10. 11591/ijece. Design of Multiplier for Medical Image Compression Using Urdhava Tiryakbhyam Sutra Suma1. Sridhar2 Vidya Vikas Institute of Engineering & Technology. Mysore. India PES College of Enginnering. Mandya. India Article Info ABSTRACT Article history: Compressing the medical images is one of the challenging areas in healthcare industry which calls for an effective design of the compression algorithms. The conventional compression algorithms used on medical images doesnAot offer enhanced computational capabilities with respect to faster processing speed and is more dependent on hardware resources. The present paper has identified the potential usage of Vedic mathematics in the form of Urdhava Tiryakbhyam sutra, which can be used for designing an efficient multiplier that can be used for enhancing the capabilities of the existing processor to generate enhance compression experience. The design of the proposed system is discussed with respect to 5 significant algorithms and the outcome of the proposed study was testified with heterogeneous samples of medical image to find that proposed system offers approximately 57% of the reduction in size without any significant loss of data. Received Oct 12, 2015 Revised Jan 5, 2016 Accepted Jan 20, 2016 Keyword: Bit loading algorithm Power allocation algorithm Resource allocation Wireless network Copyright A 2016 Institute of Advanced Engineering and Science. All rights reserved. Corresponding Author: Suma. Vidya Vikas Institute of Engineering and Technology. Mysore. India Email: sumaaldur@gmail. INTRODUCTION With the advancement of the modernized technique for diagnosis of the diseases, the healthcare system has undergone a revolutionary change by adopting medical image processing. Although medical image processing is a part of digital image processing, but it could be said as an advance applications of digital image processing as the complexities and charecteristics associated with medical images are completely different from natural images . Usually, medical images are captured using various existing image capturing devices e. X-ray. CT-scan. MRI etc. Such medical images posses valuable information which is used for mainly three purposes e. diagnosing the diseases of the subject, i. storing the images in the database for further clinical studies for medical students, and . research and development of novel ideas to mitigate the diagnosed disease. Unfortunately, the medical images . CT scan images. MRI images etc. gives information embedded in maximum resolution per slices of the data capture and thereby increasing the size of image from gigabytes to terabytes for just one medical report. However, with the data storage being cheaply available in forms of distributed servers and cloud environment, storage is never a problem for the medical images. The entire problem shoots up when the circumstances calls for transmitting the medical Such circumstance is when the healthcare industry uses telemedicine . The concept of telemedicine allows the patient and physician to interact with each other using the network resources which always has limitations of bandwidth. Such applications of telemedicine works in different principle e. doctor would like to access the medical image using downlink transmission, or i. doctor could like to perform some processing on medical images that resides on different machines along with other simultaneous data transfer. Hence, such applicability of telemedicine as well as upcoming robotic surgery calls for highly robust compression algorithm with better response rate. Here the term response will mean how fast the Journal homepage: http://iaesjournal. com/online/index. php/IJECE IJECE ISSN: 2088-8708 A 1141 algorithm can perform encoding of the streams of medical data and generate reconstructed image without losing significant information of clinical value. From past decade, there are various algorithms and techniques introduced by various researchers for carrying out medical image compression, where majority of the studies have their own advantages and disadvantages. However, it is found that existing studies are much focused on size reduction and less focus on computational capabilities e. processor speed, high resolution, slicing the components of image etc. Hence, in this regards, we came across the evolution of Vedic mathematics and its potential application that can be readily applicable in signal processing . The principle of Vedic mathematics is based on the 16 principle that are originally termed as Sutra . In order to understand the applicability of Vedic mathematics, we will need to understand that conventional compression techniques donAot focus on enhancing the processor speed owing to lack of valuable concepts of We found that Vedic mathematics offers a specific sutra for multiplier called as Urdhva Tiryakbhyam sutra, which can be highly useful for designing a new multiplier. This Vedic multiplier will be assisting is performing compression with faster speed as compared with the conventional schemes. Therefore, we make use of this multiplier in designing a processor friendly and cost effective medical compression algorithm. The prime purpose of the proposed system is to ensure reduction of computational time as well as retaining higher contents of the image contents. Section 2 discusses about the research methodology and research and discussion has been discusses in Section 3. And finally Section 4 makes some concluding remarks. Background In the recent technologies, the size of hard disks of computer and network systems has increased, but the use of medical image continues to grow exponentially at the same time require a better image and better visual quality of image. These motivate the need of compression methods. In these professional technologies, such as medical images, large amounts of dataAos are required each and every day. So image compression has becomes a necessary to ensure that their storage of data as well as transfer in insecure medium networks. The work of Urbaniak et al. basically focused on the improvement of the diagnostic important images for diagnostic purpose and to get better compression. Here the author investigated the image compression of artifacts resulting from JPEG 2000 and JPEG. The performance parameters selected for the study are PSNR, quality of image. Structural Similarity Quality Measure (SSIM), and compression ratio. This result indicates that the compression ratio based on ROC gives better visual quality as well as SSIM gives better performances. Saini et al. have performed an experiment and gives a comparison in image compression like HAAR wavelet. Bi-orthogonal Wavelets. Daubechies Wavelets and Coiflets wavelets. These algorithms are performed and testing has done for different medical images to reduce the image size and less storage requirements and it is relevant to diagnostically important regions. The outcome of the study was found with better image quality with an effective compression ratio. Nassiri et al. , done an experiment on medical image compression for diagnostically important regions using Discrete Wavelet Transforms. To minimize the total degradation and get a better compression ratio of images, the DWT compression includes hard and soft thresholding decomposition methods. The study was tested with grayscale thoracic cage image of size 512x512. The study of a volumetric diagnostically important region of medical images using 3-D listless embed block algorithm has been presented by the author Sudha et al. Here the author modified this algorithm using Set Partitioned Embedded Block Coder (SPECK) methods to get high inter band correlation. Here the MRI images are used for testing the purposes. This algorithm improves the compression ratio and as well as the improve the visual perception of images quality. John et al. have presented a high security, high transmission system for transmitting medically, diagnostically important reports for military and very high secure environment dataAos using hardware implementations methods. The authors have carried out encryption of an image on FPGA [Virtex 5 XC5VLX110T] and have also implemented a 16-stage pipeline module for achieving an encryptions rate of 35. 5 Gbit/s with 2140 configurable logic blocks. Tiwari et al. have presented multiple approaches solve a problems of storing or transmitting large number of medical data or images using different algorithms like DCT. DWT and Compressive Sensing techniques. In this paper the author performed a compressing and reconstruction technique for MRI image. CT image and Ultrasound image and the author given comparisons of all the three different medical images by using different algorithms. Here, the performance parameters taken are PSNR. MSE, compression ratio, quality of image along with storage The work by Gupta et al. have presents a method of lossless image compression for medical images using predictive coding techniques as well as integer wavelet transform based on minimum entropy This paper presents a hybrid image compression a technique that combines an integer wavelet and predictive algorithms to enhance the performance of lossless compression. The evaluation of this technique was done over greyscale medical image essentially using transformation technique along with Design of Multiplier for Medical Image Compression Using Urdhava Tiryakbhyam Sutra (Sum. A ISSN: 2088-8708 predictive-based approach. The proposed medical image compression techniques offer a higher level of compression ratio and minimum entropy when it is applied to many several levels of test images. The paper presented by Kunchigi et al. have presented a study of a Vedic mathematics approach applying for the design of 2-D DCT applications in medical image processing. Here, the author uses an Urdhva Tiryagbhyam Vedic sutra for the multiplication techniques in DCT. This paper also studied different Vedic sutras used for multiplication techniques like. Nikhilam sutra. Urdhva-Tiryagbhya sutra. The results obtained in 3 different cases, like single digit Vedic sutra, two digit Vedic sutra and finally three digit Vedic sutra. At finally, the three digit Vedic sutra gives best and better result among the three. , higher image quality as well as better image compression ratio. Sarala et al. have presented a paper on image compression using multi-level 2-D DWT as well as Vedic mathematic methods. A traditional 2-D compression algorithm consumes more power as well as more memory for image compression. But by using the Vedic mathematics algorithm used for image compression and reconstruction gives a better image compression ratio as well as better visual quality of image. This method is very useful for medical images. Vedic multiplier uses half adder and full adder. This method of 4-level 2D DWT techniques attempts to increase the image resolution. The simulations are done using Matlab 2008a version as well as ModelSim 6. version and it is implemented using Xilinx and FPGA Spartan 3 kit. Least but not last, another approach by Sowjanya et al. , implemented an approach of 2-D DCT architecture using Reversible Vedic Adder It is used to reduce the size of the image as well as in both 1-D and 2-D image processing It can also be used to calculate the 1-D and 2-D DCT using minimum number of hardware Here, the author designed a novel method by using Xilinx ISE 12. Verilog HDL tools. consumes less power consumption as well as less memory due to replacement of adder with reversible Vedic mathematics approaches. This novel algorithm improves the overall system performance in image Hence, although there are various categories of compression algorithms witnessed most recently in medical images, but very few of them are found to address the computational complexities. Therefore, there is a need of a study that could potentially address such computational issues. The Problem Usually, the size of the medical images is too large to store. Owing to the advancement in the medical studies, there are increasing dependencies associated with an efficient storage and cost effective transmission of the medical images. A medical image from Sky Scan x-ray device . generates images of size 8000 x 8000 pixels, which in nutshell leads to generation of 64 MB of image data just for one slice of the CT scan image. Therefore, other sophisticated medical imaging devices like MRI can generate around 1. GB of image data, which are not only difficult to store but highly challenging to perform transmission. Hence, such problem calls for performing compression technique. At present, various computing tools e. CUDA . Qt-Threaded . OpenCL . etc are in use in the processing such massive size of the medical images. To some extent, the discussion of the problem associated with the medical image compression was done in our prior work . The problems that have been identified for the proposed system are as follows: C Less Focus on Computational Capability: The existing technique mainly focuses on performing compression by reducing the sizes of medical images of uniform dimensions. We strongly feel that focus on medical image compression should be also extended towards ensuring the computational capability of the system from the hardware viewpoint. Hence, there is a need of the system that can provide an efficient and hardware friendly standards to perform compression of medical images. C Less Emphasis on Multipliers: Multipliers plays a critical role in digital image processing especially while performing compression. Efficient design of multiplier while performing compression always increases the speed of processor, which is extremely important in telemedicine. However, there is less work focused using conventional technique for adoption of multipliers in medical image compression. Even the usages of multipliers by using Vedic mathematics were only tested on VLSI or FPGA platforms with narrowed experimental studies. Hence, there is a need of an effective design of multiplier using cost effective computational approach to ensure processor performance while handling compression of sophisticated medical images for carrying out compression. C Less Emphasis on data redundancies: It is found that majority of the existing techniques on medical image compression ignores considering elimination of redundant data. Although majority of the existing techniques uses lossless approach for performing compression, but it seems to ignore the data redundancies that results in either less PSNR values or non-supportability in colored image compression. The above problems are the prime identification factors of the problems found after reviewing the existing system. Hence, the problem statement of the proposed study can be stated as Ae AuIt is a computationally challenging task to design a technique that ensures processor friendly as well as cost effective compression scheme on all sorts of medical images. Ay. IJECE Vol. No. June 2016 : 1140 Ae 1151 IJECE ISSN: 2088-8708 A 1143 The Proposed Solution The prime goal of the proposed system is to enhance the performance of medical image compression by applying an effective multiplier design motivated from Vedic mathematics. The design of the multiplier is done considering the algorithm called as Urdhava Tiryakbhyam sutra that is responsible for performing vertical and crosswise multiplication. Figure 1. Schema of Proposed Medical Image Compression The design of the proposed algorithm is based on a fact that addition of image elements by concurrent techniques will lead to partial products and this capability can be further strengthened by incorporating Urdhava Tiryakbhyam sutra to promote parallelism. The system allows simultaneous computation of summations of the elements being generated by the partial multiplication. Hence, such forms of multiplier do not depend on processor and its frequency of clock. In other language, the proposed approach of Vedic mathematics allows the system to execute the Vedic multiplier without any significant dependency on frequency of clock. Hence, the contribution of the proposed system can be briefed as. C To read the medical image and is applicable on both grayscale as well as colored images. C To apply multiple block-wise operation for the given image. We test it using block size of 8x8, 16x16, and 32x32. C To apply the Urdhava Tiryakbhyam sutra on each test block for performing compression of the medical C To address the redundancy problems while performing large medical image compression by using runlength coding. C To generate the encoded image and check for its quality in the respective reconstructed image using PSNR and MSE. C To check the consistency of the proposed technique on multiple type of medical images on standard RESEARCH METHOD The implementation of the proposed system is carried out using Matlab using normal 32 bit machine. The proposed system considers a medical image as an input and performs compression based on Urdhava Tiryakbhyam approach, which performs vertical and crossover multiplication in Vedic mathematics. In order to closely observe the processing time of the compression algorithms being applied, we choose to carry out this experiment on numerous machines with multiple processor of core-i5, dual core. AMD etc. In order to Design of Multiplier for Medical Image Compression Using Urdhava Tiryakbhyam Sutra (Sum. A ISSN: 2088-8708 ease off the computational complexities associated with the processing of larger size of medical images, we consider converting the input image (Iori. to grayscale. Our Algorithm-1 will show the basic step for performing this conversion that result in grayscale image (Igra. Algorithm-1: Reading Input Image Input: Input Image (Iori. Output: Grayscale Image (Igra. Start Read Iorig Resize IorigE 256 x 256. Convert the Iorig to Grayscale (Igra. End After obtaining the grayscale image (Igra. , the next step is to perform the compression technique using Vedic compression algorithm that is highlighted in Algorithm-2. The algorithm takes the input of Igray and then performs the distinct block processing of either of the size 8x8, 16x16, and 32x32. The sizes of the blocks are fed to the system using the user interface and hence are of string type. According to the Algorithm2, the size of the considered block is increased to double precision and stored in matrix R. The system will also increase the precision of Igray to double for better evaluation. The next phase of implementation will be to execute the operation of Vedic compression by applying Vedic multiplier. For this purpose, the size of the Igray is evaluated and is mapped in a separate matrix of row and column elements. The algorithm also considers certain extra zero elements to the Igray matrix and then it performs compression. In order to carry out a compression, the system designs a new function for Vedic multiplier . s shown in Line-5 of Algorithm-. , where N is size of X. T is transposition matrix, p is a matrix with element 1-(N-. and q is a matrix with element 0 to (N-. Finally, the algorithm applies the Vedic Multiplier on the squared blocks of the image for a size R. The outcome of this algorithm is a compressed version of an image using Vedic Multiplier. Algorithm-2: Vedic Compression Algorithm (VCA) Input: Grayscale Image (Igra. Output: Compressed Image using Vedic approach Start Initialize the size of the block . x8 || 16x16 || 32x. REDouble the precision of block size. XEDouble the precision of Igray. Evaluate the size of X Apply function of Vedic multiplier A A V mult ( , cos( p T . A . pi / 2 N ) Apply compression Vcomp A A . X . A T Perform block wise Vedic Compression Vcomp . lock ) CC CC X , [ R . R ] End The outcome of the Algorithm-2 is subjected to the quantization technique. However, we prefer to perform the quantization in bit discrete stage compared to common style of applying quantization in image Algorithm-3 highlights the steps being used for carrying out quantization over the compressed image . rom Algorithm-. We consider a threshold value TH=5 for analysis purpose. The initial step of this algorithm mainly evaluates the size of the compressed image i. Vcomp. and then it maps the size of it to IJECE Vol. No. June 2016 : 1140 Ae 1151 IJECE ISSN: 2088-8708 A 1145 a matrix with P . and Q. For all the elements of this matrix considering the maximum limits of P and Q elements, the proposed system attempts to compare the absolute value of compressed image i. Vcomp. with the threshold TH. Under any circumstances, if the absolute value of compressed image i. Vcomp. is found to be within the limits of threshold TH, than the system assign zero value to the compressed image and increases its count k to read other block elements. The system also initialize the quantization value to be 8 and evaluates minimum and maximum arguments of compressed for making it applicable for equation shown in Line-15 of Algorithm-3. The outcome of this algorithm is a quantized image. Algorithm-3: Quantization of Compressed Image Input: Threshold (TH). Quantization value (Q). Vcomp. Output: Quantized Image of Vcomp. Start Initialize the threshold to remove smaller values (TH=. Map the size of Vcomp. EP. FOR i=1 to P FOR j=1 to Q IF Vcomp.