یک روش مبتنی بر یادگیری چنددقتی برای تطبیق تصاویر پزشکی چندکیفیتی
محورهای موضوعی : مهندسی برق و کامپیوترسیدهسمیه آل حجت خسمخی 1 , محمدرضا کیوان‌پور 2
1 - دانشگاه آزاد اسلامی، واحد قزوین
2 - دانشگاه الزهرا
کلید واژه: تطبيق تصوير تطبیق تصویر چندکیفیتی تصویربرداری پزشکی درخت دوتایی تبدیل موجک مختلط,
چکیده مقاله :
هدف اصلی در روش‌های مختلف تطبیق تصویر، پیداکردن پارامترهای تبدیل برای نگاشت دقیق یک تصویر بر روی مختصات تصویر دیگر است. در پزشکی، برقراری ارتباط دقیق میان داده‌های تصاویر پزشکی درکاربردهایی نظیر تشخیص و درمان از اهمیت بسیاری برخوردار است و بر این اساس، روش‌های متعددی براي تطبیق تصاویر ارائه شده است. مقایسه نتایج الگوریتم‌های مختلف، انگیزه اصلی این پژوهش گردیده تا بتوان الگوریتم جدید ترکیبی ارائه و پیاده‌سازی نمود که از دقت بالایی برای تطبیق تصاویر چندکیفیتی برخوردار باشد. خودکارسازی فرایند تطبیق با بهره‌گیری از رویکرد یادگیری ماشین، نوآوری مقاله حاضر نسبت به روش‌های پیشین به شمار می‌رود. به این منظور، روش پیشنهادی به نام یادگیری چنددقتی از ترکیب یک روش تجزیه چنددقتی و یک شبکه عصبی سلسله مراتبی بهره می‌گیرد که با استفاده از ویژگی‌های سراسری تصویر، پارامترهای تبدیل را یاد گرفته و از پارامترهایِ تبدیلِ به دست آمده از فرایند یادگیری ، برای تطبیق تصاویر استفاده می‌کند. روش پیشنهادی بر روی پایگاه داده تصاویر پزشکی دانشگاه واندربیلت پیاده‌سازی و آزمون شده و نتایج به دست آمده دقت قابل قبولی را برای روش پیشنهادی در مقایسه با سایر روش‌ها نشان می‌دهد.
The main purpose in various methods of image registration is to find the transformation parameters for accurate mapping an image onto another image coordinates. In medical sciences creating a precise mapping between medical images data is very important in application such as diagnosis and treatment. Accordingly, several approaches have been proposed for image registration. The compression of results and performance between different image registration algorithms was the main motivation for this research to design and implement a new hybrid algorithm so that provide high accuracy in multimodal image registration. Automating the image registration process by using machine learning approach is the innovation of this method compared to previous ones. To this end, the proposed method which is named multi resolution learning is composed of multi resolution decomposition and a hierarchical neural network which it learn the transformation parameters by using global properties of the image and uses learned transformation parameter for image registration. The proposed method is implemented and tested on the medical images of Vanderbilt university database. Experiment result show acceptable accuracy for the proposed method compared with other methods.
[1] J. P. W. Pluim, "Mutual information based registration of medical images: a survey," IEEE Trans. on Medical Imaging, vol. 22, no. 8, pp. 986-1004, Aug. 2003.
[2] R. Suganya, K. Priyadharsini, and S. Rajaram, "Intensity based image registration by maximization of mutual information," Int. J. of Computer Application, vol. 1, no. 20, 5 pp., 2010.
[3] J. Chen and J. Tian, "Real - time multi - modal rigid registration based on a novel symmetric - SIFT descriptor," Progress in Natural Science, vol. 19, no. 5, pp. 643-651, 10 May 2009.
[4] M. R. Keyvanpour and S. Alehojat, "Analytical comparison of learning based methods to increase the accuracy and robustness of registration algorithms in medical imaging," Int. J. of Advanced Science and Technology, vol. 41, pp. 61-70, Apr. 2012.
[5] B. Zitova, J. Flusser, and F. Sroubek, "Image registration: a survey," Image and Vision Computing, vol. 21, no. 11, pp. 977-1000, 2003.
[6] G. Coppini, S. Diciotti, and G. Valli, "Matching of medical images by self - organizing neural networks," Pattern Recognition Letters, vol. 25, no. 4, pp. 341-352, Feb. 2004.
[7] S. Wang, S. Lei, and F. Chang, "Image registration based on neural network," in Proc. of the 5th Int. Conf. on Information Technology and Application in Biomedicine, pp. 74-77, 30-31 May 2008.
[8] M. Fornefett, K. Rohr, and H. S. Stiehl, "Elastic registration of medical images using radial basis functions with compact support," in Proc. IEEE Computer Society Conf. on Computer Vision and Pattern Recognition, vol. 1, pp. 1063-6919, 23-25 Jun. 1999.
[9] R. Rajeswari and A. A. Irudhayaraj, "Implementation of radial basis function for fMRI registration," Int. J. of Computer Theory and Engineering, vol. 3, no. 2, pp. 287-291, Apr. 2011.
[10] S. Oldridge, S. Fels, and G. Miller, "Classification of image registration problems using support vector machines," in Proc. of the Int. Conf. on Computer Vision, pp. 360-366, 5-7 Jan. 2011.
[11] C. Wachinger and N. Navab, "Manifold learning for multi - modal image registration," BMVC, 2010.
[12] H. –M. Chan, A. C. S. Chung, S. C. H. Yu, A. Norbash, and W. M. Wells, "Multi-modal image registration by minimizing kullback - leibler distance between expected and observed joint class histogram," in Proc. IEEE Computer Society Conf. on Computer Vision and Pattern Recognition, vol. 2, p. 570, 18-20 Jun. 2003.
[13] A. C. S. Chung, R. Gan, and W. M. Wells III, Robust Multi-Modal Image Registration Based on Prior Joint Intensity Distributions and Minimization of Kullback- Leibler Distance, HKUST CSE Technical Report, 2007.
[14] R. Liao, et al., "Learning-based 2D/3D rigid registration using Jensen-SShannon divergence for image-guided surgery," Medical Imaging and Augmented Reality, Lecture Notes in Computer Science, vol. 4091, pp. 228-235, 2006.
[15] D. Lee, et al.,, "Learning similarity measure for multi-modal 3D image registration," in Proc. IEEE Conf. on Computer Vision and Pattern Recognition, CVPR'2009, pp. 186-193, 20-25 Jun. 2009.
[16] C. Chaux, L. Duval, and J. C. Pesquet, "Image analysis using a dual-tree M - band wavelet transform," IEEE Trans. on Image Processing, vol. 15, no. 6, pp. 2397-2412, Aug. 2006.
[17] R. Mavudila, M. Cherkaoui, L. Masmoudi, and N. Hassanain, "A combined dual - tree complex wavelet (DT - CWT) and bivariate shrinkage for ultrasound medical images despeckling," Int. J. of Computer Applications, vol. 49, no. 14, pp. 42-49, Jul. 2012.
[18] A. B. Abche, F. Yaacoub, A. Maalouf, and E. Karam, "Image registration based on neural network and fourier transform," in Proc. of the 28th IEEE EMBS Annual Int. Conf., pp. 4803-4806, Aug. 2006.
[19] J. West, et al., "Comparison and evaluation of retrospective intermodality image registration techniques," SPIE Proc. on Medical Imaging: Image Processing, vol. 2710, pp. 332-347, 1996.
[20] A. Roche, G. Malandain, X. Pennec, and N. Ayache, "The correlation ratio as a new similarity measure for multimodal image registration," in Proc. 1st Int. Conf.e on Medical Image Computing and Computer-Assisted Intervention, MICCAI'98, vol. 1496, pp. 1115-1124, 1998.
[21] R. W. K. So and A. C. S. Chung, "Learning - based multi - modal rigid image registration by using Bhattacharyya distances," in Proc. 33rd Annual Int. Conf. of the IEEE on Biomedical Imaging: from Nano to Macro, pp. 368-371, Boston, Massachusetts, US, 2011.
[22] D. L. Collins and A. P. Zijdenbos, "Design and construction of a realistic digital brain phantom," IEEE Trans. Med. Img., vol. 17, no. 3, pp. 463-468, Jun. 1998.
[23] C. A. Cocosco, V. Kollokian, R. K. S. Kwan, G. Bruce Pike, and A. C. Evans, "BrainWeb: online interface to a 3D MRI simulated brain database," NeuroImage, vol. 5, no. 4, 1997.
