Detecting coronary artery disease on polar images using graph neural networks
Keywords:
Single-photon emission tomography (SPECT), Coronary artery disease, Myocardial perfusion imaging (MPI), Graph neural networksAbstract
Nowadays, myocardial perfusion imaging (MPI) with single-photon emission tomography (SPECT) is an advanced method used to detect coronary artery disease (CAD). However, analyzing and diagnosing CAD from SPECT images are a difficult, time-cosuming job and depends heavily on the experience of the doctors. In this paper, we propose a method using graph neural networks on polar maps from SPECT image to detect CAD automatically. Polar maps are transformed into graph using pixel clustering algorithm. Graphs are then fed into the network for training, extracting features and predicting the labels. The experimental models in turn give the following accuracy: GCN: 79.34%; GatedGCN 78.45%; GAT: 76.42%; MoNet: 80.1%
References
LeCun Y, Bengio Y, Hinton G, “Deep learning,” Nature, 521 (2015), pp. 436-444, doi: 10.1038/nature14539.
Shen D, Wu G, Suk H-I, “Deep learning in medical image analysis,” Annu Rev Biomed Eng, 19 (2017), pp: 221–248, doi: 10.1146/annurev-bioeng-071516-044442.
Andree Esteva, Brett Kuprel, Roberto A. Novoa, Justin Ko, Susan M. Swetter, Helen M. Blau & Sebastian Thrun, “Derma-tologist-level classification of skin cancer with deep neural networks,” Nature, 542 (2017), pp: 115–118, doi: 10.1038/nature21056.01.
Julian Betancur et. al., “Deep Learning for Prediction of Obstructive Disease from Fast Myocardial Perfusion SPECT,” JACC: Cardiovascular Imaging, vol. 11(11) (2018), pp. 1654-1663.
Duvenaud, David, et al, "Convolutional networks on graphs for learning molecular fingerprints," arXiv preprint arXiv:1509.09292 (2015).
Cranmer, Miles D., et al, "Learning symbolic physics with graph networks," arXiv preprint arXiv:1909.05862 (2019).
Kipf, Thomas N., and Max Welling, "Semi-supervised classification with graph convolutional networks," arXiv preprint arXiv:1609.02907 (2016).
Monti F, Boscaini D, Masci J, Rodola E, Svoboda J, Bronstein MM, “Geometric deep learning on graphs and manifolds using mixture model CNNs,” In: 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE (2017).
Veličković, Petar, et al, "Graph attention networks," arXiv preprint arXiv:1710.10903 (2017).
Bresson, Xavier, and Thomas Laurent, "Residual gated graph convnets," arXiv preprint arXiv:1711.07553 (2017).
Dwivedi, Vijay Prakash, et al, "Benchmarking graph neural networks," arXiv preprint arXiv:2003.00982 (2020).
Achanta R, Shaji A, Smith K, Lucchi A, Fua P, Süsstrunk S, “SLIC superpixels compared to state-ofthe-art superpixel methods,” IEEE Trans Pattern Anal Mach Intell, (2012) ;34(11):2274-2282.
