Automated Deep Learning Analysis of Angiography Video Sequences for Coronary Artery Disease

The evaluation of obstructions (stenosis) in coronary arteries is currently done by a physician’s visual assessment of coronary angiography video sequences. It is laborious, and can be susceptible to interobserver variation. Prior studies have attempted to automate this process, but few have demonstrated an integrated suite of algorithms for the end-to-end analysis of angiograms. We report an automated analysis pipeline based on deep learning to rapidly and objectively assess coronary angiograms, highlight coronary vessels of interest, and quantify potential stenosis. We propose a 3-stage automated analysis method consisting of key frame extraction, vessel segmentation, and stenosis measurement. We combined powerful deep learning approaches such as ResNet and U-Net with traditional image processing and geometrical analysis. We trained and tested our algorithms on the Left Anterior Oblique (LAO) view of the right coronary artery (RCA) using anonymized angiograms obtained from a tertiary cardiac institution, then tested the generalizability of our technique to the Right Anterior Oblique (RAO) view. We demonstrated an overall improvement on previous work, with key frame extraction top-5 precision of 98.4%, vessel segmentation F1-Score of 0.891 and stenosis measurement 20.7% Type I Error rate.

[1]  D. Katritsis,et al.  Limitations of coronary angiography: An underestimated problem? , 1991, Clinical cardiology.

[2]  Da Young Lee,et al.  Automatic stenosis recognition from coronary angiography using convolutional neural networks , 2020, Comput. Methods Programs Biomed..

[3]  Uri Shaham,et al.  Automated Characterization of Stenosis in Invasive Coronary Angiography Images with Convolutional Neural Networks , 2018, ArXiv.

[4]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[5]  Krystian Mikolajczyk,et al.  Deep Segmentation and Registration in X-Ray Angiography Video , 2018, BMVC.

[6]  Ajmal S. Mian,et al.  Deep Keyframe Detection in Human Action Videos , 2018, ArXiv.

[7]  Emmanuelle Gouillart,et al.  scikit-image: image processing in Python , 2014, PeerJ.

[8]  Hamid Soltanian-Zadeh,et al.  Local feature fitting active contour for segmenting vessels in angiograms , 2014, IET Comput. Vis..

[9]  Geoffrey E. Hinton,et al.  Lookahead Optimizer: k steps forward, 1 step back , 2019, NeurIPS.

[10]  Akihiro Sato,et al.  Quantitative vascular measurements in arterial occlusive disease. , 2005, Radiographics : a review publication of the Radiological Society of North America, Inc.

[11]  June-Goo Lee,et al.  Angiography‐Based Machine Learning for Predicting Fractional Flow Reserve in Intermediate Coronary Artery Lesions , 2019, Journal of the American Heart Association : Cardiovascular and Cerebrovascular Disease.

[12]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[13]  D. Bravata,et al.  Carotid artery stenosis: wide variability in reporting formats--a review of 127 Veterans Affairs medical centers. , 2013, Radiology.

[14]  Kilian Q. Weinberger,et al.  Densely Connected Convolutional Networks , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[15]  Su Yang,et al.  Deep learning segmentation of major vessels in X-ray coronary angiography , 2019, Scientific Reports.

[16]  D. Holmes,et al.  Estimation of the effects of angioplasty on coronary stenosis using quantitative video angiography. , 1985, Catheterization and cardiovascular diagnosis.

[17]  David Beymer,et al.  Automatic Selection of Keyframes from Angiogram Videos , 2010, 2010 20th International Conference on Pattern Recognition.

[18]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[19]  R. Popp,et al.  Accuracy and reproducibility of visual coronary stenosis estimates using information from multiple observers , 1992, Clinical cardiology.

[20]  Danni Ai,et al.  Multichannel Fully Convolutional Network for Coronary Artery Segmentation in X-Ray Angiograms , 2018, IEEE Access.

[21]  Mohammad H. Jafari,et al.  Segmentation of vessels in angiograms using convolutional neural networks , 2018, Biomed. Signal Process. Control..

[22]  Yannis Avrithis,et al.  Label Propagation for Deep Semi-Supervised Learning , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[23]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[24]  S. Kelsey,et al.  Reliability of assessing change with sequential coronary angiography. , 1982, American heart journal.

[25]  Liyuan Liu,et al.  On the Variance of the Adaptive Learning Rate and Beyond , 2019, ICLR.

[26]  Ching Y. Suen,et al.  A fast parallel algorithm for thinning digital patterns , 1984, CACM.

[27]  S. M. Collins,et al.  Visual estimates of percent diameter coronary stenosis: "a battered gold standard". , 1988, Journal of the American College of Cardiology.

[28]  David Beymer,et al.  Automatic detection of coronary stenosis in X-ray angiography through spatio-temporal tracking , 2014, 2014 IEEE 11th International Symposium on Biomedical Imaging (ISBI).

[29]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[30]  Zengchang Qin,et al.  Automated identification and grading of coronary artery stenoses with X-ray angiography , 2018, Comput. Methods Programs Biomed..