ناحیه‌بندی بطن چپ در تصاویر اکوکاردیوگرافی با استفاده از یادگیری منیفلد و تلفيق میدان برداری جهت‌دار دینامیکی
محورهای موضوعی : مهندسی برق و کامپیوترنجمه مشهدی 1 , حمید بهنام 2 , احمد شالباف 3 , زهرا علیزاده ثانی 4
1 - دانشگاه علم و صنعت ایران
2 - دانشگاه علم و صنعت ایران
3 - دانشگاه علم و صنعت ایران
4 - مرکز درمانی، آموزشی و تحقیقاتی قلب و عروق شهید رجایی
کلید واژه: کانتور فعال بطن چپ یادگیری منیفلد الگوریتم نگاشت محلي خطي اکوکاردیوگرافی,
چکیده مقاله :
بیماری‌های قلبی شايع‌ترين علت مرگ و مير در جهان هستند. بررسی عملکرد بطن چپ که وظيفه خون‌رساني به تمامي نقاط بدن را دارد، در تشخیص بیماری‌های قلبی بسیار حائز اهمیت است. تعیین و ردیابی خودکار مرزهای ديواره بطن چپ در طول یک سیکل قلبی جهت كمي‌سازي عملکرد ديواره بطن چپ قلبي به جهت تشخيص بيماري‌هاي مختلف قلبي از جمله بيماري ايسکمي استفاده مي‌شود. در این مقاله، روش خودکار جديدي برای تعیین مرز ديواره بطن چپ در تصاوير اکوکاردیوگرافی يک سيکل قلبي ارائه شده که در اين الگوريتم از ترکيب روش‌هاي کانتور فعال هندسی بر اساس نیروی خارجی تلفیق میدان برداری جهت‌دار و يادگيري منيفلد استفاده شده است. در اين روش، ابتدا تصاوير اکوکارديوگرافي يک سيکل قلبي با استفاده از يکي از پرکاربردترين روش‌هاي يادگيري منيفلد به نام نگاشت محلي خطي به فضاي دوبعدي نگاشت مي‌شود. در اين فضاي ويژگي جديد ارتباط بين فريم‌هاي يک سيکل قلبي به خوبي نشان داده مي‌شود. سپس تعيين مرز ديواره بطن چپ در طول یک سيکل قلبي با استفاده از روش کانتور فعال هندسی بر اساس نیروی خارجی تلفیق میدان برداری جهت‌دار انجام می‌گیرد. در این روش مرز نهایی یک فریم به عنوان مرز اولیه فریم بعدی در نظر گرفته شده و به منظور افزايش دقت تعيين مرز ديواره بطن چپ و همچنين جلوگيري از انحراف مرز، میزان حرکت مجاز مرز ناشي از روش کانتور فعال هندسي از ارتباط بین فریم‌ها، متناظر با فریم جاری و فریم قبلی، در فضای دوبعدی محدود مي‌گردد. برای ارزیابی کمی روش پیشنهادی از 9 توالی تصاویر اکوکاردیوگرافی (5 داوطلب سالم و 4 بیمار) استفاده شده است. مرز ديواره بطن چپ به دست آمده با روش پیشنهادی با مرز ديواره به دست آمده توسط پزشک متخصص باتجربه (استاندارد طلایی) مقایسه شده و نتايج به دست آمده حاکي از دقت بالاي روش پيشنهادي در تعيين مرز ديواره بطن چپ مي‌باشد.
Cardiac diseases are the major causes of death throughout the world. The study of left ventricular (LV) function is very important in the diagnosis of heart diseases. Automatic tracking of the boundaries of the LV wall during a cardiac cycle is used for quantification of LV myocardial function in order to diagnose various heart diseases including ischemic disease. In this paper, a new automatic method for segmentation of the LV in echocardiography images of one cardiac cycle by combination of manifold learning and active contour based dynamic directed vector field convolution (DDVFC) is proposed. In this method, first echocardiography images of one cardiac cycle have been embedded in a two dimensional (2-D) space using one of the most popular manifold learning algorithms named Locally Linear Embeddings. In this new space, relationship between these images is well represented. Then, segmentation of the LV wall during a cardiac cycle is done using active contour based DDVFC. In this method, final contour of each segmented frame is used as the initial contour of the next frame. In addition, in order to increase the accuracy of the LV segmentation and also prevent the boundary distortion, maximum range of the active contour motion is limited by Euclidean distances between consequent frames in resultant 2-D manifold. To quantitatively evaluate the proposed method, echoacardiography images of 5 healthy volunteers and 4 patients are used. The results obtained by our method are quantitatively compared to those obtained manually by the highly experienced echocardiographer (gold standard) which depicts the high accuracy of the presented method.
[1] http://www.cardiosmart.org/HeartDisease.
[2] A. M. Katz, Physiology of the Heart, 4th ed. Philadelphia, Pa: Lippincott Williams & Wilkins, 2006.
[3] http://www.heart.org/HEARTORG/Conditions/HeartAttack/PreventionTreatmentofHeartAttack
[4] http://www.sciencedaily.com/articles/i/ischaemic_heart_disease.htm.
[5] J. J. Soraghan and S. K. Setarehdan, "Automatic echocardiographical feature extraction for left ventricular wall motion and volume changes visualization," in Proc. of the IEEE 23rd Annual Int. Conf. of the Engineering in Medicine and Biology Society, vol. 2, pp. 1653-1656, 2001.
[6] L. Han and W. Qi - Sheng, "Motion object tracking algorithm using an improved geometric active contour model," in Proc. of 3rd Int. Congress on Image and Signal Processing, CISP, vol. 1, pp. 331-334, 16-18 Oct. 2010.
[7] X. Hang, N. L. Greenberg, and J. D. Thomas, "A geometric deformable model for echocardiographic image segmentation," Computers in Cardiology, vol. 29, pp. 77-80, 2002.
[8] X. Min, X. Shunren, and W. Shiwei, "Geometric active contour model with color and intensity priors for medical image segmentation," in Proc. IEEE - EMBS 27th Annual Int. Conf. Engineering in Medicine and Biology Society, pp. 6496-6499, Shanghai, China, 17-18 Jan. 2006.
[9] G. Raghotham Reddy et al., "Fast global region based minimization of satellite and medical imagery with geometric active contour and level set evolution on noisy images," Recent Advances in Intelligent Computational Systems, RAICS, IEEE, pp. 696-700, 22-24 Sep 2011.
[10] A. Shalbaf et al., "Automatic detection of end systole and end diastole within a sequence of 2 - D echocardiographic images using modified Isomap algorithm," in Proc. 1st Middle East Conf. on Biomedical Engineering, MECBME,. pp. 217-220, 21-24 Feb. 2011.
[11] A. Sarti, C. Corsi, E. Mazzini, and C. Lamberti, "Maximum likelihood segmentation of ultrasound images with Raleigh distribution," in Proc. IEEE Conf. on Computer Cardiology, pp. 329-332, 19-22 Sep. 2004.
[12] Q. Duan, E. Angelini, S. Homma, and A. Laine, "Tracking endocardium using optical flow along iso-value curve," in Proc. IEEE Int. Conf. of the on Engineering in Medicine and Biology Society, vol. 1, pp. 707-710, 2006.
[13] J. Liang and Y. Wang, "Improved GVF based left ventricle segmentation from cardiac MR images using radial B - snake model," in Proc. IEEE Int. Conf. on Young Computer Scientists, pp. 1000-1005, 18-21 Nov. 2008.
[14] L. Bing and S. T. Acton, "Vector field convolution for image segmentation using snakes," in Proc. IEEE Int. Conf. on Image Processing, pp. 1637-1640, 8-11 Oct. 2006.
[15] S. Ganbari Maman, et al., "Fully automatic segmentation of left ventricle in a sequence of echocardiography images of one cardiac cycle by dynamic directional vector field convolution (DDVFC) method and manifold learning," Biomedical Engineering: Applications, Basis and Communications, vol. 25, no. 2, 15 pp., Apr. 2012.
[16] Z. Xingfu and R. Xiangmin, "Two dimensional principal component analysis based independent component analysis for face recognition," in Proc. Int Conf. on Multimedia Technology, ICMT, pp. 934-936, 26-28 Jul. 2011.
[17] H. Moeinzadeh et al., "Combination of harmony search and linear discriminate analysis to improve classification," in Proc. 3rd Asia Int. Conf. on Modelling & Simulation, AMS'09, pp. 131-135, 25-29 May 2009.
[18] J. B. Tenenbaum V. de Silva, and J. Langford, "A global geometric framework for nonlinear dimensionality reduction," Science, vol. 290, no. 5500, pp. 2319-2323, 22 Dec. 2000.
[19] S. T. Roweis and L. K. Saul, "Nonlinear dimensionality reduction by locally linear embedding," Springer, vol. 290, no. 5500, pp. 2323-2326, 22 Dec. 2000.
[20] L. K. Saul and S. T. Roweis, "Think globally, fit locally: unsupervised learning of low dimensional manifolds," J. of Machine Learning Research, vol. 4, no. 1, pp. 119-155, Jan. 2003.
[21] M. Belkin and P. Niyogi, "Laplacian eigenmaps and spectral techniques for embedding and clustering," Neural Information Processing Systems, vol. 14, pp. 585-591, 2001.
[22] O. Kouropteva, O. Okun, and M. Pietikainen, "Incremental locally linear embedding algorithm," in Proc. 14th Scandinavian Conf. on Image Analysis, SCIA'05, pp. 521-530, 2005.
[23] D. Donoho and C. Grimes, "Hessian eigenmaps: new locally linear embedding techniques for high-dimensional data," in Proc. of the National Academy of Sciences, vol. 102, no. 21, pp. 7426-7431, 2005.
[24] X. Huo, X. Ni, and A. K. Smith, "A survey of manifold-based learning methods," Mining of Enterprise Data, Ch. 1, pp. 691-745, 2007.
[25] http://www.asecho.org.
