臺灣博碩士論文加值系統

對於醫生在診斷心血管疾病測量超音波心電圖心臟大小和功能評估是一個很重要的評判依據，包含左心室區域以計算體積和射血分數。然而，由於超音波影像品質和散斑噪聲，左心室的邊界難以界定。手動定位腔室邊界的每個點需要相當長的時間。因此，我們設計了一個系統，自動對不同的心臟超音波影像進行分類，分割左心室來以分析體積，以幫助醫生發現可能與心血管疾病相關的症狀。在本文中，我們提出了一種分類算法，該算法利用骨架化作為每個心臟超音波的特徵，分類的平均準確率高達 90%。此外，再利用多級閾值下效果使用掃描算法來尋找左心室腔邊緣，Dice相似係數平均值在86.3%左右。

Measuring the echocardiogram assessment of cardiac size and function is decisive for inferring cardiovascular disease, bounding out the area of the Left ventricle in order to calculate volume and ejection fraction. However, the boundary of the Left ventricle is hard to define because of the image quality and speckle noise. Manually locating every point of the chambers’ boundary needs a considerable amount of time. Therefore, we designed a system that automatic classifies different views, segmenting the Left ventricle cavity for the purpose of analyzing the volume, to assist doctors on finding symptoms that might be related to cardiovascular disease. In this paper, we propose a classifying algorithm that utilizes skeletonize as the feature to every single view, which the average Accuracy is up to 90%. In addition, the scanning algorithm works well with multi-level threshold finding the edge of the Left ventricle cavity, and the average of Dice similarity coefficient is around 86.3%.

ABSTRACT
摘要
List of Figures
List of Tables
I. Introduction
II. Workflow
III. Proposed Approach
3.1. Skeletonize
3.2. Chamber Center
3.3. Multilevel threshold
3.4. Scanning
3.5. Regression
3.6. Cardiac cycle
IV. Results
4.1. View Classification
4.2. Image Segmentation of Left Ventricle
4.3. Dice similarity coefficient
V. Discussion
VI. Conclusions
VII. References

[1]Roberto M. Lang, Luigi P. Badano, Victor Mor-Avi, Jonathan Afilalo, Anderson Armstrong, Laura Ernande, Frank A. Flachskampf, Elyse Foster, Steven A. Goldstein, Tatiana Kuznetsova, Patrizio Lancellotti, Denisa Muraru, Michael H. Picard, Ernst R. Rietzschel, Lawrence Rudski, Kirk T. Spencer, Wendy Tsang, Jens-Uwe Voigt, "Recommendations for Cardiac Chamber Quantification by Echocardiography in Adults: An Update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging", Journal of the American Society of Echocardiography, 2015, Pages 1-39.e14, https://doi.org/10.1016/j.echo.2014.10.003.
[2]Xiaohong Gao, Wei Li, Martin Loomes, Lianyi Wang, A fused deep learning architecture for viewpoint classification of echocardiography, Information Fusion, 2017, Pages 103-113, doi: 10.1016/ j.inffus. 2016.11.007.
[3]Hanan Khamis, Grigoriy Zurakhov, Vered Azar, Adi Raz, Zvi Friedman, Dan Adam, "Automatic apical view classification of echocardiograms using a discriminative learning dictionary", Medical Image Analysis, 2017, pp.15-21, doi :10.1016/j.media.2016.10.007.
[4]Madani, A., Arnaout, R., Mofrad, M. et al. "Fast and accurate view classification of echocardiograms using deep learning". npj Digital Medice 1, 2018, Article number: 6, https://doi.org/10.1038/s41746-017-0013-1
[5]S. Ebadollahi, Shih-Fu Chang and H. Wu, "Automatic view recognition in echocardiogram videos using parts-based representation", Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004, pp. II-II, doi: 10.1109/CVPR.2004.1315137.
[6]Bosch JG, Mitchell SC, Lelieveldt BP, Nijland F, Kamp O, Sonka M, Reiber JH. "Automatic segmentation of echocardiographic sequences by active appearance motion models", IEEE Transactions on Medical Imaging, 2002, pp. 1374-1383, doi: 10.1109/TMI.2002.806427. PMID: 12575874.
[7]Wu, Xuanyan & Li, Xiuling & Mou, Gang & Wang, Dechun & He, Yan & Li, Zhengdong, "Identification of the left ventricle endocardial border on two-dimensional ultrasound images using the convolutional neural network Unet", 2018 Ural Symposium on Biomedical Engineering, Radioelectronics and Information Technology (USBEREIT), 2018, pp. 76-78, doi: 10.1109/USBEREIT.2018.8384554.
[8]Ozan Oktay, Enzo Ferrante, Konstantinos Kamnitsas, "Anatomically Constrained Neural Networks (ACNNs): Application to Cardiac Image Enhancement and Segmentation", IEEE Transactions on Medical Imaging, 2018, pp. 384-395, doi: 10.1109/TMI.2017.2743464.
[9]Shakiba Moradi, Mostafa Ghelich Oghli, Azin Alizadehasl, Isaac Shiri, Niki Oveisi, Mehrdad Oveisi, Majid Maleki, Jan Dhooge, "MFP-Unet: A novel deep learning based approach for left ventricle segmentation in echocardiography", Physica Medica, 2019, Pages 58-69, doi.org 10.1016 j.ejmp.2019.10.001.
[10]Yasser Ali, Farrokh Janabi-Sharifi, Soosan Beheshti, "Echocardiographic image segmentation using deep Res-U network", Biomedical Signal Processing and Control, 2021, https:// doi.org/ 10.1016/ j.bspc. 2020.102248.
[11]Waleed Abu-Ain, Siti Norul Huda Sheikh Abdullah, Bilal Bataineh, Tarik Abu-Ain, Khairuddin Omar, "Skeletonization Algorithm for Binary Images", Procedia Technology, 2013, Pages 704-709, https:// doi.org/ 10.1016/j.protcy.2013.12.248.
[12]Tarik A.H. Abu-Ain, Waleed A.H. Abu-Ain, Siti Norul Huda Sheikh Abdullah and Khairuddin Omar, "Off-line Arabic Character-Based Writer Identification – a Survey", International Journal on Advanced Science, Engineering and Information Technology, 2011, pp.161-166., doi.org/10.18517/ijaseit.1.2.35.
[13]Gopakumar, Rajesh, N. V. Subbareddy, Krishnamoorthi Makkithaya and Dinesh U. Acharya. “Script Identification from Multilingual Indian Documents using Structural Features”, Journal of Computing, 2010, pp. 106-111
[14]Mohamed A. Ali, "An efficient thinning algorithm for arabic ocr systems", Signal & Image Processing An International Journal (SIPIJ), 2012, pp.31-38, doi:10.5121/sipij.2012.3303
[15]Na, Shi, Liu Xumin and Guan. Yong, "Research on k-means Clustering Algorithm: An Improved k-means Clustering Algorithm", 2010 Third International Symposium on Intelligent Information Technology and Security Informatics, 2010, pp. 63-67, doi: 10.1109/IITSI.2010.74.
[16]Yau-Hwang Kuo, Chang-Shing Lee, Chao-Lieh Chen, "High-stability AWFM filter for signal restoration and its hardware design", Fuzzy Sets and Systems, 2000, pp. 185-202, doi:10.1016/S0165-0114(98)00112-2.
[17]Scikit-learn: Machine Learning in Python, Pedregosa et al., JMLR 12, 2011, pp. 2825-2830.
[18]Raafs AG, Boscutti A, Henkens MTHM, van den Broek WWA, Verdonschot JAJ, Weerts J, Stolfo D, Nuzzi V, Manca P, Hazebroek MR, Knackstedt C, Merlo M, Heymans SRB, Sinagra G. "Global Longitudinal Strain is Incremental to Left Ventricular Ejection Fraction for the Prediction of Outcome in Optimally Treated Dilated Cardiomyopathy Patients", Journal of the American Heart Association, 2022, doi: 10.1161/JAHA.121.024505.
[19]Karlsen, S., Dahlslett, T., Grenne, B. et al. "Global longitudinal strain is a more reproducible measure of left ventricular function than ejection fraction regardless of echocardiographic training", Cardiovascular ultrasound, 2019, doi: 10.1186/ s12947-019-0168-9
[20]Oglat AA, Matjafri MZ, Suardi N, Oqlat MA, Abdelrahman MA, Oqlat AA. "A Review of Medical Doppler Ultrasonography of Blood Flow in General and Especially in Common Carotid Artery", Journal of medical ultrasound, 2018, pp.3-13, doi: 10.4103/JMU.JMU_11_17.