Motion Estimation in Ultrasonic Images
Motion analysis for soft tissue in ultrasonic images is a new and important research field both for diagnostic applications and image processing. This procedure is helpful in identifying pathological abnormalities in ultrasonic motion images, which have rarely been studied due to the challenge presented by the speckle de-correlation problem. In the traditional motion estimation methods, a rigid body motion is usually assumed, which is not suitable for analyzing the motion of soft tissues. Hence, a motion estimation system designed for soft tissue motion is in great demand.
A feature metamorphosis based motion estimation algorithm for tracking soft tissue motion in ultrasonic image sequences was presented. In this algorithm, a metamorphosis-based method generates the dense flow field and an energy-based refinement process is employed to modify the morph flows. Based on the system structure and the basic constraints adopted in the proposed method, a process suitable for soft tissue motion estimation was produced. Compared with other systems designed for soft tissue motion in ultrasonic images, the proposed method is much simpler and is both accurate and efficient. The potential applications for ultrasonic images motion estimation includes studies on muscle elasticity, tumor detection, investigation of vascular tissue properties and other areas of clinical investigation.
According to the rapid developments in data acquisition equipment, an extension from 2D image data to 3D volume data is inevitable. 3D flow estimation will demand further development.
This feature metamorphosis based algorithm can be easily extended into 3D flow estimation. This extension surely will include more motion information and provide more clinical evidence for better diagnoses. One of the major tasks is improving our method for better motion estimation and more realistic applicability in the future.
Motion Estimation in Tagged MR Images
In this paper, we have presented an automatic algorithm for tag identification and tracking from tagged MR images. By applying a series of tag extracting steps, tag edge images are generated as references for estimating initial displacement fields, which provide suggestions for tag line propagation. The predefined energies confine the tag lines according to the tag properties and force the tag lines to move toward its optimal positions.
In the traditional algorithms, the tags are usually defined as the low brightness regions that are easily affected by noises and the tag decay problem. In the proposed method, the tags can be defined more precisely by extracting the tag signals. By preprocessing the tagged MR images with the morphological operations, the tag signals can be extracted and defined. The tag decay problem can be eliminated since the gradient definition is much less sensitive to the brightness decay.
The MR tagging schemes use a set of RF pulses to place the tag patterns perpendicular to the imaging slice and form two orthogonal sets of regular tag lines. The number of available tag lines reflects the spatial resolution of the tag pattern that usually influences the accuracy of the estimated motion. However, increasing the spatial resolution of the tag pattern implies the decreasing of the tag spacing, and small tag spacing could accompany ambiguities in tag tracking if the tag displacement is larger than half of the tag spacing.
By analyzing the spatio-temporal patterns, the temporal information can provide preliminary motion fields, which place the tracked tag line close to the correct one. Hence, it reduces the matching ambiguities between neighboring tag lines since the tag patterns are periodical and are all alike. To refine the tracked tag lines, the spatial information is then added to provide the image properties that obtained from tagging. It helps the tag lines to be lain along its real positions.
Briefly, the temporal information provides reliability and the spatial information provides accuracy.
In the tag tracking process, both the temporal and spatial information are adopted to deform the tag lines. By integrating the spatial and temporal information from the tagged MR image sequences, the tag lines can be tracked well. The deformation fields are then reconstructed from the tracked results for myocardium contractile ability analysis. Finally, strain maps are generated for infarction inspection where spatial and temporal deformed information are both shown in a 2D image.
The current system is capable of analyzing 2D tagged MR images and achieves good results.
It has been experimented with both the motion synthesis and in vivo data sets. Both the reconstructed displacement and deformation fields show sufficient accuracy for measuring myocardial deformations from the tagged MR image sequences.
References
[1] A. Hein, and W. D. O’Brien, Jr., “Current Time-Domain Methods for Assessing Tissue Motion by Analysis from Reflected Ultrasound Echoes–A Review,” IEEE Trans. on Ultrasonics, Ferroelectics, and Frequency Control, Vol. 40, No. 2, pp. 84-102, 1993.
[2] R. L. Maurice, and M. Bertrand, “Lagrangian Speckle Model and Tissue-Motion Estimation-Theory,” IEEE Trans. on Medical Imaging, Vol. 18, No. 7, pp. 593-603, 1999.
[3] M. G. Strintzis, and I. Kokkinidis, “Maximum Likelihood Motion Estimation in Ultrasound Image Sequences,” IEEE Signal Processing Letters, Vol. 4, No. 6, pp. 156-157, 1997.
[4] F. Yeung, S. F. Levinson, D. Fu, and K. J. Parker, “Feature-Adaptive Motion Tracking of Ultrasound Image Sequences Using A Deformable Mesh,” IEEE Trans. on Medical Imaging, Vol. 17, No. 6, pp. 945-956,1998.
[5] F. Yeung, S. F. Levinson, and K. J. Parker, “Multilevel and Motion Model-Based Ultrasonic Speckle Tracking Algorithms,” Ultrasound in Medicine & Biology, Vol. 24, No. 3, pp.
427-441, 1998.
[6] G. E. Mailloux, F. Langlois, and P. Y. Simard, “Restoration of the Velocity Field of the Heart from Two-Dimensional Echocardiograms,” IEEE Trans. on Medical Imaging, Vol. 8, No. 2, pp. 143-153, 1989.
[7] S. J. Ericson, MD, “High-Resolution Imaging of the Musculoskeletal System,” Radiology, Vol. 205, No. 3, Dec 1997.
[8] B. K. P. Horn, and B. G. Schunck, “Determining Optical Flow,” Artificial Intelligence, Vol.
17, pp. 185-203, 1981.
[9] H. S. Wang, and R. M. Mersereau, “Fast Algorithms for the Estimation of Motion Vectors,”
IEEE Trans. on Image Processing, Vol. 8, No. 3, pp. 435-438, 1999.
[10] H. Gu, Y. Shirai, and M. Asada, “MDL-Based Segmentation and Motion Modeling in a Long Image Sequence of Scene with Multiple Independently Moving Objects,” IEEE Trans.
on Pattern Analysis and Machine Intelligence, Vol. 18, No. 1, pp. 58-64, 1996.
[11] D. G. Sim, O. K. Kwon, and R. H. Park, “Object Matching Algorithms Using Robust Hausdorff Distance Measures,” IEEE Trans. on Image Processing, Vol. 8, No. 3, pp.
425-429, 1999.
[12] K. Y. Wohn, J. Wu, and R. W. Brockett, “A Contour-Based Recovery of Image Flow:
Iterative Transformation Method,” IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol. 13, No. 8, pp. 746-760,1991.
[13] C. Kervrann, and F. Heitz, “Statistical Deformation Model-Based Segmentation of Image Motion,” IEEE Trans. on Image Processing, Vol. 8, No. 4, pp. 583-588, 1999.
[14] E. Mémin, and P. Pérez, “Dense Estimation and Object-Based Segmentation of the Optical Flow with Robust Techniques,” IEEE Trans. on Image Processing, Vol. 7, No. 5, pp.
703-719, 1998.
[15] C. H. Lee, and L. H. Chen, “A Fast Motion Estimation Algorithm Based on the Block Sum Pyramid,” IEEE Trans. on Image Processing, Vol. 6, No. 11, pp. 1587-1590, 1997.
[16] J. Y. Lu, K. S. Wu, and J. C. Lin, “Fast Full Search in Motion Estimation by Hierarchical Use of Minkowski’s Inequality (HUMI),” Pattern Recognition, Vol.31, No. 7, pp. 945-952, 1998.
[17] M. Kass, A. Witkin, and D. Terzopoulos, “Snacks: Active Contour Models,” International Journal of Computer Vision, Vol. 1, No. 4, pp. 321-331, 1988.
[18] F. Leymarie and M. D. Levine, “Tracking Deformable Objects in the Plane Using an Active Contour Model,” IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol. 15, No. 6, pp. 617-634, 1993.
[19] Y. Xu, J. B. Weaver, D. M. Healy, Jr., and J. Lu, “Wavelet Transform Domain Filters: A Spatially Selective Noise Filtration Technique,” IEEE Trans. on Image Processing, Vol. 3, No. 6, pp. 747-758, 1994.
[20] D. Ruprecht, and H. Müller, “Image Warping with Scattered Data Interpolation,” IEEE Computer Graphics and Applications, Vol. 15, No. 2, pp. 37-43, 1995.
[21] J. Gomes, L. Darsa, B. Costa, and L. Velho, Warping and Morphing of Graphical Objects, Morgan Kaufmann Publishers, 1999, pp.234-235.
[22] T. Y. Lee, Y. C. Lin, L. Lin, Y. N. Sun, “Fast Feature-based Metamorphosis and Operator Design,” EuroGraphics'98, Journal of Computer Graphics Forum, Vol. 17, No. 3, pp.
C15-22, 1998.
[23] T. Beier, and S. Neely, “ Feature-Based Image Metamorphosis,” Computer Graphics (Proceedings of SIGGRAPH 92), Vol. 26, No. 2, pp. 35-42 , 1992.
[24] J. L. Barron. D. J. Fleet, S. S. Beauchemin, “Performance of Optical Flow Techniques,”
International Journal of Computer Vision, Vol. 12, No. 1, pp.43-77, 1994.
[25] H. Liu, T. H. Hong, M. Herman, T. Camus, and R. Chellapa, “Accuracy vs Efficiency Trade-offs in Optical Flow Algorithms,” Computer Vision and Image Understanding, Vol.
72, No. 3, pp. 271-286, 1998.
[26] M. A. Guttman, E. A. Zerhouni, and E. R. McVeigh, “Analysis of Cardiac Function from MR Images,” IEEE Computer Graphics and Applications, vol. 17, no. 1, pp. 30-38, 1997.
[27] J. M. Bundy, and C. H. Lorenz, “TAGASIST: A Post-Processing and Analysis Tools Package for Tagged Magnetic Resonance Imaging,” Computerized Medical Imaging and Graphics, vol. 21, no. 4, pp. 225-232, 1997.
[28] M. Stuber, M. B. Scheidegger, S. E. Fisher, E. Nagel, F. Steinemann, O. M. Hess, P.
Boesiger, “Alterations in the Local Myocardial Motion Pattern in Patients Suffering From Pressure Overload Due to Aortic Stenosis,” Circulation, vol. 100, pp. 361-368, 1999.
[29] C. M. Kramer, C. J. McCreery, L. Semonik, W. J. Rogers, T. P. Power, A. Shaffer, and N.
Reichek, “Global Alterations in Mechanical Function in Healed Reperfused First Anterior Myocardial Infarction,” Journal of Cardiovascular Magnetic Resonance, vol. 2, no. 1, pp.33-41, 2000.
[30] H. B. Hillenbrand, J. A. C. Lima, D. A. Bluemke, G. M. Beache, and E. R. McVeigh,
“Assessment of Myocardial Systolic Function by Tagged Magnetic Resonance Imaging,”
Journal of Cardiovascular Magnetic Resonance, vol. 2, no. 1, pp. 57-66, 2000.
[31] J. Bogaert, H. Bosmans, A. Maes, P. Suetens, G. Marchal, and F. E. Rademakers, “Remote Myocardial Dysfunction After Acute Anterior Myocardial Infarction: Impact of Left Ventricle Shape on Regional Function,” Journal of the American College of Cardiology, vol.
35, no. 6, pp. 1525-1534, 2000.
[32] E. A. Zerhouni, D. M. Parish, W. J. Rogers, A. Yang, and E. P. Shapiro, “Human Heart:
Tagging with MR Imaging – A Method for Noninvasive Assessment of Myocardial Motion,” Radiology, vol. 169, no. 1, pp. 59-63, 1988.
[33] L. Axel, and L. Dougherty, “MR Imaging of Motion with Spatial Modulation of Magnetization,” Radiology, vol. 171, no. 3, pp. 841-845, 1989.
[34] B. D. Bolster, E. R. McVeigh, and E. A. Zerhouni, “Myocardial Tagging in Polar Coordinates with Use of Striped Tags,” Radiology, vol. 177, no. 3, pp. 769-772, 1990.
[35] B. K. P. Horn, and B. G. Schunck, “Determining Optical Flow,” Artificial Intelligence, vol.
17, pp. 185-203, 1981.
[36] J. L. Prince, and E. R. McVeigh, “Motion Estimation from Tagged MR Image Sequences,”
IEEE Trans. on Medical Imaging, vol. 11, no. 2, pp. 238-249, 1992.
[37] L. Dougherty, J. C. Asmuth, A. S. Blom, L. Axel, and R. Kumar, “Validation of an Optical Flow Method for Tag Displacement Estimation,” IEEE Trans. on Medical Imaging, vol. 18, no. 4, pp. 359-363, 1999.
[38] M. A. Guttman, J. L. Prince, and E. R. McVeigh, “Tag and Contour Detection in Tagged MR Images of the Left Ventricle,” IEEE Trans. on Medical Imaging, vol. 13, no. 1, pp. 74-88, 1994.
[39] E. Atalar, and E. R. McVeigh, “Optimization of Tag Thickness for Measuring Position with Magnetic Resonance Imaging,” IEEE Trans. on Medical Imaging, vol. 13, no. 1, pp.
152-160, 1994.
[40] T. S. Denney Jr., “Estimation and Detection of Myocardial Tags in MR Image without User-Defined Myocardial Contours,” IEEE Trans. on Medical Imaging, vol. 18, no. 4, pp.
330-344, 1999.
[41] S. Kumar, and D. Goldof, “Automatic Tracking of SPAMM Grid and the Estimation Parameters from Cardiac MR Images,” IEEE Trans. on Medical Imaging, vol. 13, no. 1, pp.
122-132, 1994.
[42] A. A. Young, D. L. Kraitchman, L. Dougherty, and L. Axel, “Tracking and Finite Element Analysis of Strip Deformation in Magnetic Resonance Tagging,” IEEE Trans. on Medical Imaging, vol. 14, no. 3, pp. 413-421, 1995.
[43] P. Radeva, A. A. Amini, and J. Huang, “Deformable B-Solids and Implicit Snakes for 3D Localization and Tracking of SPAMM MRI Data,” Computer Vision and Image Understanding, vol. 66, no. 2, pp. 163-178, 1997.
[44] A. A. Amini, Y. Chen, R. W. Curwen, V. Mani, and J. Sun, “Coupled B-Snake Grids and Constrained Thin-Plate Splines for Analysis of 2-D Tissue Deformations from Tagged MRI,” IEEE Trans. on Medical Imaging, vol. 17, no. 3, pp. 334-356, 1998.
[45] J. Huang, D. Abenschein, V. G. Dávila-Román, and A. A. Amini, “Spatio-Temporal Tracking of Myocardial Deformations with a 4-D B-Spline Model from Tagged MRI,” IEEE Trans.
on Medical Imaging, vol. 18, no. 10, pp. 957-972, 1999.
[46] C. Ozturk, and E. R. McVeigh, “Four Dimensional B-Spline Based Motion Analysis of Tagged Cardiac MR Images,” Proceedings of SPIE Medical Imaging, 1999.
[47] A. A. Amini, Y. Chen, M. Elayyadi, and P. Radeva, “Tag Surface Reconstruction and Tracking of Myocardial Beads from SPAMM-MRI with Parametric B-Spline Surfaces,”
IEEE Trans. on Medical Imaging, vol. 20, no. 2, pp. 94-103, 2001.
[48] W. G. O'Dell, E. R. McVeigh, C. C. Moore, E. A. Zerhouni, “Implementation of Displacement Field Fitting for Calculating 3D Myocardial Deformations from Parallel-Tagged MR Images,” Proceedings of the 16th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, pp. 551-552, 1994.
[49] T. S. Denney Jr., and J. L. Prince, “3D Displacement Field Reconstruction on an Irregular Domain from Planar Tagged Cardiac MR Images,” Proceedings of the 1994 IEEE Workshop on Motion of Non-Rigid and Articulated Objects, pp. 172 -177, 1994.
[50] J. Park, D. Metaxas, and L. Axel, “Analysis of Left Ventricular Wall Motion Based on Volumetric Deformable Models and MRI-SPAMM,” Medical Image Analysis, vol. 1, no. 1, pp. 53-71, 1996.
[51] P. Clarysse, C. Basset, L. Khouas, P. Croisille, D. Friboulet, C. Odet, I. E. Magnin,
“Two-Dimensional Spatial and Temporal Displacement and Deformation Field Fitting from Cardiac Magnetic Resonance Tagging,” Medical Image Analysis, vol. 4, pp. 253-268, 2000.
[52] M. Kass, A. Witkin, and D. Terzopoulos, “Snakes: Active Contour Models,” International Journal of Computer Vision, vol. 1, no. 4, pp. 321-331, 1987.
[53] C.-W. Ngo, T.-C. Pong, and H.-J. Zhang, “Motion Analysis and Segmentation Through Spatiao-Temporal Slices Processing,” IEEE Trans. on Image Processing, vol. 12, no. 3, pp.
341-355, 2003.
[54] J. Gomes, L. Darsa, B. Costa, and L. Velho, Warping and Morphing of Graphical Objects, Morgan Kaufmann Publishers, pp.234-235, 1999.
[55] T. Arts, W. Hunter, A. Douglas, A. Muijtjens, and R. Reneman, “Description of the Deformation of the Left Ventricle by a Kinematic Model,” Journal of Biomechanics, vol. 25, no. 10, pp. 1119-1127, 1992.
[56] E. Waks, J. L. Prince, and A. S. Douglas, “Cardiac Motion Simulator for Tagged MRI,”
Proceedings of the IEEE Workshop on Mathematical Methods in Biomedical Image Analysis, pp. 182-191, 1996.
[57] J. L. Barron., D. J. Fleet, and S. S. Beauchemin, “Performance of Optical Flow Techniques,”
International Journal of Computer Vision, vol. 12, no. 1, pp. 43-77, 1994.
[58] D. S. Fieno, R. J. Kim, E. L. Chen, J. W. Lomasney, F. J. Klocke, and R. M. Judd,
“Contrast-Enhanced Magnetic Resonance Imaging of Myocardium at Risk: Distinction Between Reversible and Irreversible Injury Throughout Infarct Healing,” Journal of the American College of Cardiology, vol. 36, no. 6, pp. 1985-1991, 2000.
[59] C. M. Kramer, W. J. Rogers, S. Mankad, T. M. Theobald, D. L. Pakstis, and Y. L. Hu,
“Contractile Reserve and Contrast Uptake Pattern by Magnetic Resonance Imaging and Functional Recovery After Reperfused Myocardial Infarction,” Journal of the American College of Cardiology, vol. 36, no. 6, pp. 1835-1840, 2000.
計畫成果自評
本計畫之運動分析系統，我們分別針對超音波影像以及磁振造影網格影像加以處理，
主要使用個人電腦直接分析醫學影像序列以獲取運動分析的定量數據，針對病變區提出客 觀警訊作為醫師診療的參考，直接有益病灶之診治。
在超音波影像方面，由於超音波取像的即時特性，我們首先針對超音波影像來進行運 動分析。但是由於超音波影像在產生時，容易受到環境因素的影響，因此容易產生 speckle 的不連續性。此研究內容針對超音波影像特質，提出利用特徵形變的方法偵測組織運動情 形，藉由運動分析的方式，偵測出組織運動的異常現象。由實驗結果我們可發現傳統的運 動計算方式，由於受到超音波雜訊的影響，使得運算偵測的結果混亂而不正確。而加入運 動流場平滑度的限制，利用特徵形變的方式，則可使得計算結果較符合真實情況之運動情 形。
在磁振造影影像方面，我們利用網格影像提供觀察組織運動或形變之非侵入性功能。
在醫療應用方面，於過去十年中，磁振造影網格影像被大量用於心臟影像運動分析及其相 關臨床應用。而由於磁振造影網格影像會隨時間之經過而自行訊號衰減，故容易造成網格
追蹤之困難。此研究內容針對MR Tag 影像特質，提出利用利用邊線追蹤的運動偵測方式，
以偵測核磁共振網格之形變，進而量測與分析心臟組織的運動。在實驗結果中，透過與利 用延遲加強的磁振造影取像技巧所得到的影像加以比較，我們可看出利用本研究所提標記 追蹤的方法，確實能達到不錯的效果。而將我們分析出的運動流場，經過應變分析後，我 們可計算出心肌收縮的形變場，經由類似牛眼圖的對應，我們可將序列中的形變場對應至 一個二維的空間時間應變圖中，藉此可偵測出心肌中異常區域的位置。利用此一標記追蹤 的方式，除了針對病患分析其病變區域及程度，此一量化數據也可作為手術後復原情形的 評估。除此之外，對於可能有心臟疾病的患者，也可藉由此一分析，了解是否已有病變徵 兆，並用來預防心肌梗塞的發生。
本研究內容符合原計劃內容，亦達成預期目標。而根據實驗確可達到不錯的結果，除 可應用於醫療臨床分析，在學術研究上亦有發表於科學期刊的價值。本期研究成果已在研 討會發表並投稿於國際期刊，在不久之將來將會發表刊出。