Predicting Strokes in Carotid Atherosclerosis Using
Transcription
Predicting Strokes in Carotid Atherosclerosis Using
ISSN(Online): 2320-9801 ISSN (Print): 2320-9798 International Journal of Innovative Research in Computer and Communication Engineering (An ISO 3297: 2007 Certified Organization) Vol. 3, Special Issue 2, March 2015 Predicting Strokes in Carotid Atherosclerosis Using Ultrasound Image Analysis A.Syed Musthaba1, N.Karthick2, Saranya.M3 Gnanamani College of Technology, Namakkal, India1,2,3 ABSTRACT: This project deals with serious and frequent cerebrovascular disease. The most common cause of stroke according to a specific diagnostic algorithm can be prevented by treating carotid atherosclerosis. This algorithm is insufficient and poses a significant research challenge to students and study in detail to find a solution. Previous solutions illustrate on this matter the potential of carotid ultrasound image analysis with a high cost and innovative imaging on plaque composition and stability. This project illustrates the potential of carotid ultrasound image analysis towards this direction with ultrasound imaging being a low cost image based features revealing the information on plaque composition and stability. Systematic applications are used to diagnose and find results and orient future research. A bright respective for clinical scenario is diagnosed for atherosclerotic patients in low cost. The overall project developed in MATLAB platform. Finally validation of results shows both in hardware and software, the data of vulnerability or non vulnerability of plaque is displayed . KEYWORDS: Ultrasoundimage, atherosclerosis, cerebrovascular, Carotid artery, Plaque. I. INTRODUCTION Stroke is a global health problem. It is the second commonest cause of death and fourth leading cause of disability worldwide. Approximately 20 million people each year will suffer from stroke and of these 5 million will not survive. In developed countries, stroke is the first leading cause for disability, second leading cause of dementia and third leading cause of death. Strokes can be classified into two major categories: ischemic and hemorrhagic. Ischemic strokes are those that are caused by interruption of the blood supply, while hemorrhagic strokes are the ones which result from rupture of a blood vesselor an abnormal vascular structure. About 87% of strokes are caused by ischemia, and the remainder by hemorrhage. Some hemorrhages develop inside areas of ischemia ("hemorrhagic transformation"). Ischemic strokes caused by artery stenosis, account for approximately 75% of all strokes. Stroke is also a predisposing factor for epilepsy, falls and depression in developed countries and is a leading cause of functional impairments, with 20% of survivors requiring institutional care after 3 months and 15% - 30% being permanently disabled .Computer aided diagnosis(CAD) of carotid atherosclerosis into symptomatic orasymptomatic is useful in the analysis of cardiac health.In the existing method only plaque characterization in carotid ultrasound scan was analyzed. In the proposed method we are going to analyses the blood clot in the carotid artery as well as going to measure the variations in the length and breadth of the carotid artery. In the existing system only experienced physicians or vascular ultrasonographers can analyses the plaque. But in the proposed system all can identify the symptoms. II.BLOCK DIAGRAM Ultrasound input image is collected and imaged is preprocessed ,algorithms are used for feature extraction, neutral network is used for training approximately 20 images using back propagation technique and images are classified into normal and abnormal images. The following block diagram describes the procedure of finding the solution given below. Copyright to IJIRCCE www.ijircce.com 57 ISSN(Online): 2320-9801 ISSN (Print): 2320-9798 International Journal of Innovative Research in Computer and Communication Engineering (An ISO 3297: 2007 Certified Organization) Vol. 3, Special Issue 2, March 2015 Figure.1 Block diagram of procedure technique. III.METHODOLOGY A.ULTRASOUND INPUT IMAGE Ultrasound imaging holds a prominent position in the diagnosis of carotid atherosclerosis. It has a number of advantages including non invasiveness, widespread availability short examination time. Pre-processing operations also known filtration is use a small neighbourhood of a pixel in an input image to get a value in the output image. Pre processing is a set of techniques whose aim is improving the input image to make it easier to process by the segmentation step. The blocked carotid artery area is zoomed and cropped. The nature of the disease is focused on the vessel wall that specifically changes the morphology of the lumen–intima interface from slow gradual lipid formation and maturing into hard plaque or loose island of hemorrhage.The following image describes the output image of finding the solution fig.9 C.DISCRETE WAVELET TRANSFORM Wavelets are extract information from many different kinds of data including audio and images. The transform generates sub band LL,LH,HL,HH each with one forth. Most of the energy is concentrated in HH, which represents the high-resolution version of the original image. A DWT is any wavelet transform for which the wavelets are discretely sampled. For 2-D signals, the 2-D DWT can be used. focuses on wavelet packets (WPs) for images. These images are represented as an m × n gray scale matrix I[i, j] where each element of the matrix represents the intensity of one pixel.These eight neighbours can be used to traverse through the matrix. The direction with which the matrix is traversed just inverts the sequence of pixels, and the 2-D DWT coefficients are the same. Figure .2 Discrete wavelet transform(2D-DWT) operation The four possible directions, which are known as decomposition corresponding to 0◦ (horizontal, Dh), 90◦ (vertical, Dv) and 45◦ or 135◦ (diagonal, Dd) orientations. The first-level 2-D DWT yields four result matrices, namely,Dh1, Dv1, Dd1 and A1, whose elements are intensity values. The sub band coding can be repeated for further decomposition. Copyright to IJIRCCE www.ijircce.com 58 ISSN(Online): 2320-9801 ISSN (Print): 2320-9798 International Journal of Innovative Research in Computer and Communication Engineering (An ISO 3297: 2007 Certified Organization) Vol. 3, Special Issue 2, March 2015 G.BIORTHOGONAL WAVELET A biorthogonal wavelet is a wavelet where the associated wavelet transform is invertible. Designing biorthogonal wavelets allows more degrees of freedom than orthogonal wavelets. One additional degree of freedom is the possibility to construct symmetric wavelet functions. Two sets of wavelets generate subspaces respectively. The basis is orthogonal; the two MRAs are said to be biorthogonal to each other. The bi-orthogonal wavelet system designed shows improvement common- used least square method in output image quality. The above four techniques of discrete wavelet transform is used to decompose the preprocessed image. By comparing the output image of four techniques it’s found that biorthoganal gives high resolution and accurate image which is given as the input image to neural network classifier. .The following image describes the output image of finding the solution fig.5 IV. FEATURE EXTRACTION Feature extraction is a special form of dimensionality reduction. The input data to an algorithm is suspected to benotoriously redundant, and then the input data will be transformed into a reduced representation set of features. Feature extraction transforms the input data into the set of features. The features set will extract the relevant information from the input data in order to perform the desired task using this reduced representation. Feature extraction simplifies a large set of data accurately. One of the major problems stems from the number of variables involved.Featureextraction methodologies analyze objects and images to extract the most prominent features that are representative of the various classes of objects. Feature extraction is a general term for methods of constructing combinations of the variables to get around these problems while still describing the data with sufficient accuracy.The following features were extracted in decomposed image: Average, Energy, Entropy, Covariance, Standard deviation. V.NEURALNETWORK CLASSIFIER(BPN TECHNIQUE) Neural networks have emerged as an important tool for classification and promising alternative to various conventional classification methods. The advantage of neural networks lies in the following theoretical aspects. Figure.3 two-layer neural network The internal weights of the neural network are adjusted according to the transactions used in the learning process. The neural network receives in addition the expected output in each transaction. This allows the modification of the weights. Thetrained neural network is used to classify new images. Neural network classifier is used to compare the trained image and input image based on score value. If the classifier output is 1.it will be the normal image, else it will be the abnormal image which indicates the presence of symptoms. Kernel configurations of neural network in classification results can be compared with neural network classifier. neural classifier performed better than other classifiers. Copyright to IJIRCCE www.ijircce.com 59 ISSN(Online): 2320-9801 ISSN (Print): 2320-9798 International Journal of Innovative Research in Computer and Communication Engineering (An ISO 3297: 2007 Certified Organization) Vol. 3, Special Issue 2, March 2015 VI.CLASSIFICATION OF NEURAL NETWORK(BPN TECHNIQUE ) The features present in the date are analyzed using the input data and an accurate description is analyzed in classification of plaque. Unknown class descriptions are classified using this model. Unclassified cases are assigned a class label in classification. Processing techniques apply to the values in a digital yield or remotely sensed scene to group pixels with similar digital number values into feature classes or categories. Classification is one of the most active research and application. Figure.4 Output imageof Biorthoganal Preprocessing VII.INSTRUMENTATION DETAILS Mat lab software is used to find the results and the results are transferred to RS-232. RS-232 is a standard for serial binary data interconnection between a DTE and a DCE. Electrical signal characteristics such as voltage levels, signaling rate, timing and slew-rate of signals, voltage with stand level, short-circuit behavior, maximum stray capacitance and cable length. Interface mechanical characteristics, pluggable connectors and pin identification. Functions of each circuit in the interface connector. Standard subsets of interface circuits for selected telecom applications. Many modern devices can exceed this speed (38,400 and 57,600 bit/s being common, and 115,200 and 230,400 bit/s making occasional appearances) while still using RS-232 compatible signal levels is suitable. A typical serial port includes specialized driver and receiver integrated circuits to convert between internal logic levels and RS232 compatible signal are sent to ZigBee transmitter and from there the signals are received in ZigBee receiver.The microcontroller that has been used for this project is from PIC series.LCD is display the result. The hardware device consists of PC output unit, microcontroller unit and ZigBee transmitter and receiver unit, LCD display, driver, alarm. The Fig.11 shows the block diagram of the device. Fig.5. Overview of instrumentation block diagram Copyright to IJIRCCE www.ijircce.com 60 ISSN(Online): 2320-9801 ISSN (Print): 2320-9798 International Journal of Innovative Research in Computer and Communication Engineering (An ISO 3297: 2007 Certified Organization) Vol. 3, Special Issue 2, March 2015 VIII. ZIGBEE TRANSMITTER ZigBee is a specification for a suite of high level communication protocols used to create personal area networks built from small, low-power digital radios.ZigBee devices can transmit data over long distances by passing data through intermediate devices to reach more distant ones, creating a mesh network; i.e., a network with no centralized control or high-power transmitter/receiver able to reach all of the networked devices. ZigBee is used in applications that require only a low data rate, long battery life, and secure networking that requires short-range wireless transfer of data at relatively low rates. Figure.6 ZigBee Transmitter IX.ZIGBEE RECEPTION The microcontroller that has been used for this project is from PIC series. PIC microcontroller is the first RISC based microcontroller fabricated in CMOS (complementary metal oxide semiconductor) that uses separate bus for instruction and data allowing simultaneous access of program and data memory.Various microcontrollers offer different kinds of memories. EEPROM, EPROM, FLASH etc. are some of the memories of which FLASH is the most recently developed. Technology that is used in PIC 16F877 is flash technology, so that data is retained even when the power is switched off. Easy Programming and Erasing are other features of PIC 16F877. X.MICRO CONTROLLER Microcontroller is converted into four wire signal of ZigBee to two wire communication channel or serial communication system. It controls the high voltage of the circuit.. Microcontrollers can run into millions of units per application. At these volumes of the microcontrollers is a commodity items and must be optimized so that cost is at a minimum. A micro controller unit (MCU) uses the microprocessor as its central processing unit (CPU) and incorporates memory, timing reference, I/O peripherals, etc on the same chip. Limited computational capabilities and enhanced I/O are special features Figure.7 ZigBee Receiver Copyright to IJIRCCE www.ijircce.com 61 ISSN(Online): 2320-9801 ISSN (Print): 2320-9798 International Journal of Innovative Research in Computer and Communication Engineering (An ISO 3297: 2007 Certified Organization) Vol. 3, Special Issue 2, March 2015 XI.DISPLAY AND ALARM When the LCD is in the off state, light rays are rotated by the two polarizer and the liquid crystaland hence the LCD appears transparent. When sufficient voltage is applied to the electrodes, the liquid crystal molecules would be aligned in a specific direction.Finally validation of results shows in LCD connector , the data of vulnerability or non vulnerability of plaque is displayed. XII. LCD DISPLAY RESULT Figure .8 LCD display result CONCLUSION AND FUTURE WORK Plaque and Blood clot identification from ultrasound images is low cost and can be used by all kind of patients. Physicians or vascular ultrasonographers detect these differences during ultrasound scans. The proposed CAD system overcomes some of the problems faced by the physicians. It can be used as a diagnostic tool in modern clinical practice since it can provide decision support with regard to carotid plaque treatment. The system is able to diagnose the of plaque and blood clot formations automatically. The Neural network algorithm is used to classify symptomatic and asymptomatic plaques and blood clot. This accuracy is relatively higher than those recorded in similar studies in the literature. The proposed technique is an effective image mining techniquewhich can serve as an efficient adjunct tool for the vascular surgeons in selecting patients for risky stenosis treatments. The accuracy of the proposed system can be improved further to be incorporated into routine clinical work flow. The future work includes studying more feature extraction techniques in order to improve the accuracy and usage. Findings ways for the motion of carotid artery using arterial elasticity. Finding measures for alteration of carotid artery in the presence of atherosclerosis. Finding solution for Blood pressure which induce compressive stress in radial directions and tensile stress. REFERENCES [1] M. L. Grønholdt., B. G. Nordestgaard, T. V. Schroeder, S. Vorst up, and H. Sillesen,June [2001]“Ultrasonic echo lucent carotid plaques predict future strokes,”Circulation,vol. 104, no. 1, pp. 68 73 [2] J. E. Wilhjelm, M. L. M. Grønholdt, B. Wiebe, S. K. Jespersen, L.K. Hansen, H. Sillesen,[Dec. 1998]“Quantitative analysis of ultrasound Bmode images of carotid atherosclerotic plaque: correlation with visual classification and histological examination,” IEEE Trans.Med. Imag., vol. 17, no. 6, pp. 910-922, [3] C. I. Christodoulou, C. S. Pattichis, M. Pantziaris, and A. N.Nicolaides, [Jul 2003] “Texture-based classification of atherosclerotic carotid plaques,”.IEEE Trans. Med. Imaging, vol. 22, no. 7, pp. 902-912. [4]AbuRahma, A. and Crotty, B (2002), “Carotid plaque ultrasonic heterogeneity and severity of stenosis”, IEEE Transactions on Medical Imaging, Vol. 22, No.7, pp. 1772-1775 [5]AimiliaGastounioti., Konstantina, S. and SpyrettaGolemati, (2013), “Toward Novel Noninvasive and Low-Cost Markers for Predicting Strokes in Asymptomatic Carotid Atherosclerosis: The Role of Ultrasound Image Analysis”, IEEE transactions on Biomedical Engineering, Vol. 60, No.14, pp. 35-40. [6]Alpers, C., Gordon, D., Strandness, D. E. and Hatsukami, T. S, (1997), “Carotid plaque morphology and clinical events”, IEEE transactions on Biomedical Engineering, Vol. 40, No.19, pp. 28-30.[7]M. L. Grønholdt., B. G. Nordestgaard, T. V. Schroeder, S.Vorstrup, and H. Sillesen, “Ultrasonic echo lucent carotid plaques predict future strokes,” Circulation, vol. 104, no. 1, pp. 68-73, 2001. Copyright to IJIRCCE www.ijircce.com 62 ISSN(Online): 2320-9801 ISSN (Print): 2320-9798 International Journal of Innovative Research in Computer and Communication Engineering (An ISO 3297: 2007 Certified Organization) Vol. 3, Special Issue 2, March 2015 BIOGRAPHY SYED MUSTHABA.A is a Assistant Professor in the Electronics and Communication Engineering Department, Gnanamani college of Technology. He received Master of Engineering (ME) degree in 2011 from Karpagamuniversity,Coimbatore,India. His research interests are Computer Networks, Image analysis. KARTHICK.N is a Assistant Professor in the Electronics and Communication Engineering Department, Gnanamani college of Technology. He received Master of Engineering (ME) degree in 2013 from Karpagam university, Coimbatore,India. His research interests areLoe power VLSI, Image analysis.. SARANYA.M is a PG Scholor in the Applied Electronics, Gnanamani college of Technology. Her research interests are Computer Networks, Image analysis, Neuralnetwork, biomedical. Copyright to IJIRCCE www.ijircce.com 63