Real-time Detection of Aortic Valve in Echocardiography using Convolutional Neural Networks

Author(s): Muhammad Hanif Ahmad Nizar, Chow Khuen Chan, Azira Khalil, Ahmad Khairuddin Mohamed Yusof, Khin Wee Lai*

Journal Name: Current Medical Imaging
Formerly: Current Medical Imaging Reviews

Volume 16 , Issue 5 , 2020


Become EABM
Become Reviewer
Call for Editor

Graphical Abstract:


Abstract:

Background: Valvular heart disease is a serious disease leading to mortality and increasing medical care cost. The aortic valve is the most common valve affected by this disease. Doctors rely on echocardiogram for diagnosing and evaluating valvular heart disease. However, the images from echocardiogram are poor in comparison to Computerized Tomography and Magnetic Resonance Imaging scan. This study proposes the development of Convolutional Neural Networks (CNN) that can function optimally during a live echocardiographic examination for detection of the aortic valve. An automated detection system in an echocardiogram will improve the accuracy of medical diagnosis and can provide further medical analysis from the resulting detection.

Methods: Two detection architectures, Single Shot Multibox Detector (SSD) and Faster Regional based Convolutional Neural Network (R-CNN) with various feature extractors were trained on echocardiography images from 33 patients. Thereafter, the models were tested on 10 echocardiography videos.

Results: Faster R-CNN Inception v2 had shown the highest accuracy (98.6%) followed closely by SSD Mobilenet v2. In terms of speed, SSD Mobilenet v2 resulted in a loss of 46.81% in framesper- second (fps) during real-time detection but managed to perform better than the other neural network models. Additionally, SSD Mobilenet v2 used the least amount of Graphic Processing Unit (GPU) but the Central Processing Unit (CPU) usage was relatively similar throughout all models.

Conclusion: Our findings provide a foundation for implementing a convolutional detection system to echocardiography for medical purposes.

Keywords: Aortic valve, heart valve, echocardiography, cardiology, convolutional neural network, deep learning.

[1]
Mozaffarian D, Benjamin EJ, Go AS, et al. Writing group members; American Heart Association Statistics committee; Stroke statistics subcommittee. Heart disease and stroke statistics-2016 update: A report from the American Heart Association. Circulation 2016; 133(4): e38-e360.
[http://dx.doi.org/10.1161/CIR.0000000000000350] [PMID: 26673558]
[2]
Moore M, Chen J, Mallow PJ, Rizzo JA. The direct health-care burden of valvular heart disease: evidence from US national survey data. Clin Outcomes Res CEOR 2016; 188: 613-27.
[3]
Rodés-Cabau J, Taramasso M, O’Gara PT. Diagnosis and treatment of tricuspid valve disease: current and future perspectives. Lancet 2016; 388(10058): 2431-42.
[http://dx.doi.org/10.1016/S0140-6736(16)00740-6] [PMID: 27048553]
[4]
Messika-Zeitoun D, Lloyd G. Aortic valve stenosis: Evaluation and management of patients with discordant grading. E-J Cardiol Prac 2018; p. 15.
[5]
Khalil A, Faisal A, Lai KW, Ng SC, Liew YM. 2D to 3D fusion of echocardiography and cardiac CT for TAVR and TAVI image guidance. Med Biol Eng Comput 2017; 55(8): 1317-26.
[http://dx.doi.org/10.1007/s11517-016-1594-6] [PMID: 27830464]
[6]
Zhang S, Zhu D, Wan Z, Cao Y. Utility of point-of-care echocardiogram in the rapid diagnosis of hypertrophic cardiomyopathy. Am J Emerg Med 2013; 31(8): 1280-2.
[http://dx.doi.org/10.1016/j.ajem.2013.04.036] [PMID: 23759682]
[7]
Grau V, Becher H, Noble JA. Registration of multiview real-time 3-D echocardiographic sequences. IEEE Trans Med Imaging 2007; 26(9): 1154-65.
[http://dx.doi.org/10.1109/TMI.2007.903568] [PMID: 17896589]
[8]
Khalil A, Faisal A, Ng SC, Liew YM, Lai KW. Multimodality registration of two-dimensional echocardiography and cardiac CT for mitral valve diagnosis and surgical planning. J Med Imaging (Bellingham) 2017; 4(3) 037001
[http://dx.doi.org/10.1117/1.JMI.4.3.037001] [PMID: 28840172]
[9]
Chai HY, Wee LK, Swee TT, Salleh ShH, Chea LY. An artifacts removal post-processing for epiphyseal region-of-interest (EROI) localization in automated bone age assessment (BAA). Biomed Eng Online 2011; 10(1): 87.
[http://dx.doi.org/10.1186/1475-925X-10-87] [PMID: 21952080]
[10]
Faisal A, Ng S-C, Goh S-L, George J, Supriyanto E, Lai KW. Multiple LREK active contours for knee meniscus ultrasound image segmentation. IEEE Trans Med Imaging 2015; 34(10): 2162-71.
[http://dx.doi.org/10.1109/TMI.2015.2425144] [PMID: 25910057]
[11]
Hossain MB, Lai KW, Pingguan-Murphy B, Hum YC, Salim MIM, Liew YM. Contrast enhancement of ultrasound imaging of the knee joint cartilage for early detection of knee osteoarthritis. Biomed Signal Process Control 2014; 13: 157-67.
[http://dx.doi.org/10.1016/j.bspc.2014.04.008]
[12]
Wee LK, Chai HY, Supriyanto E. Surface rendering of three dimensional ultrasound images using VTK. J Sci Ind Res 2011; 70(6): 421-6.
[13]
Reményi B, Wilson N, Steer A, et al. World Heart Federation criteria for echocardiographic diagnosis of rheumatic heart disease--an evidence-based guideline. Nat Rev Cardiol 2012; 9(5): 297-309.
[http://dx.doi.org/10.1038/nrcardio.2012.7] [PMID: 22371105]
[14]
Krizhevsky A, Sutskever I, Hinton GE. Imagenet classification with deep convolutional neural networks. Adv Neur Inform Process Syst 2012; 25(2): 1097-5.
[15]
Szegedy C, Liu W, Jia Y, et al. Going deeper with convolutions. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Boston, MA, USA New Jersey: IEEE 2015; pp. 1-9.
[16]
Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition 2014. arXiv Prepr arXiv14091556.
[17]
Lin T-Y, Maire M, Belongie S, et al. Microsoft coco: Common objects in context. In: European conference on computer vision. Zurich, Switzerland. Berlin: Springer 2014; pp. 740-55.
[18]
Everingham M, Van Gool L, Williams CKI, Winn J, Zisserman A. The pascal visual object classes (voc) challenge. Int J Comput Vis 2010; 88(2): 303-38.
[http://dx.doi.org/10.1007/s11263-009-0275-4]
[19]
Tzutalin LabelImg 2015. [cited 2018 June 20]. Available from:. https://github.com/tzutalin/labelImg
[20]
Bradski G. The OpenCV Library Dr Dobb’s J Softw Tools. 2000.
[21]
Huang J, Rathod V, Sun C, et al. Speed/accuracy trade-offs for modern convolutional object detectors. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, HI, USA, New Jersey: IEEE 2017; pp. 3296-7.
[http://dx.doi.org/10.1109/CVPR.2017.351]
[22]
Duruöz CI, Ozcelik T, Shimizu Y. Digital video decoding, buffering and frame-rate converting method and apparatus Google Patents. 2003.
[23]
Richard P, Birebent G, Coiffet P, Burdea G, Gomez D, Langrana N. Effect of frame rate and force feedback on virtual object manipulation. Presence Teleoperators Virtual Environ 1996; 5(1): 95-108.
[http://dx.doi.org/10.1162/pres.1996.5.1.95]
[24]
Claypool M, Claypool K, Damaa F. The effects of frame rate and resolution on users playing first person shooter games Multimedia Computing and Networking 2006 International Society for Optics and Photonics; San Jose, CA, USA. Bellingham: SPIE 2006; p. 607101.
[http://dx.doi.org/10.1117/12.648609]


open access plus

Rights & PermissionsPrintExport Cite as

Article Details

VOLUME: 16
ISSUE: 5
Year: 2020
Published on: 28 May, 2020
Page: [584 - 591]
Pages: 8
DOI: 10.2174/1573405615666190114151255

Article Metrics

PDF: 15
HTML: 4
EPUB: 1
PRC: 1