LIFE: A Generalizable Autodidactic Pipeline for 3D OCT-A Vessel Segmentation

被引:12
作者
Hu, Dewei [1 ]
Cui, Can [1 ]
Li, Hao [1 ]
Larson, Kathleen E. [2 ]
Tao, Yuankai K. [2 ]
Oguz, Ipek [1 ]
机构
[1] Vanderbilt Univ, Dept Elect Engn & Comp Sci, 221 Kirkland Hall, Nashville, TN 37235 USA
[2] Vanderbilt Univ, Dept Biomed Engn, Nashville, TN 37235 USA
来源
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2021, PT I | 2021年 / 12901卷
关键词
OCT angiography; Self-supervised; Vessel segmentation; ANGIOGRAPHY; EVOLUTION; SCALE;
D O I
10.1007/978-3-030-87193-2_49
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Optical coherence tomography (OCT) is a non-invasive imaging technique widely used for ophthalmology. It can be extended to OCT angiography (OCT-A), which reveals the retinal vasculature with improved contrast. Recent deep learning algorithms produced promising vascular segmentation results; however, 3D retinal vessel segmentation remains difficult due to the lack of manually annotated training data. We propose a learning-based method that is only supervised by a self-synthesized modality named local intensity fusion (LIF). LIF is a capillary-enhanced volume computed directly from the input OCT-A. We then construct the local intensity fusion encoder (LIFE) to map a given OCT-A volume and its LIF counterpart to a shared latent space. The latent space of LIFE has the same dimensions as the input data and it contains features common to both modalities. By binarizing this latent space, we obtain a volumetric vessel segmentation. Our method is evaluated in a human fovea OCT-A and three zebrafish OCT-A volumes with manual labels. It yields a Dice score of 0.7736 on human data and 0.8594 +/- 0.0275 on zebrafish data, a dramatic improvement over existing unsupervised algorithms.
引用
收藏
页码:514 / 524
页数:11
相关论文
共 33 条
[1]   Recurrent residual U-Net for medical image segmentation [J].
Alom, Md Zahangir ;
Yakopcic, Chris ;
Hasan, Mahmudul ;
Taha, Tarek M. ;
Asari, Vijayan K. .
JOURNAL OF MEDICAL IMAGING, 2019, 6 (01)
[2]   Initialization, noise, singularities, and scale in height ridge traversal for tubular object centerline extraction [J].
Aylward, SR ;
Bullitt, E .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 2002, 21 (02) :61-75
[3]   A texture-based 3D region growing approach for segmentation of ICA through the skull base in CTA [J].
Bozkurt, Ferhat ;
Kose, Cemal ;
Sari, Ahmet .
MULTIMEDIA TOOLS AND APPLICATIONS, 2020, 79 (43-44) :33253-33278
[4]   Application of OCT- angiography to characterise the evolution of chorioretinal lesions in acute posterior multifocal placoid pigment epitheliopathy [J].
Burke, T. R. ;
Chu, C. J. ;
Salvatore, S. ;
Bailey, C. ;
Dick, A. D. ;
Lee, R. W. J. ;
Ross, A. H. ;
Carreno, E. .
EYE, 2017, 31 (10) :1399-1408
[5]   A Deep Learning Approach to Denoise Optical Coherence Tomography Images of the Optic Nerve Head [J].
Devalla, Sripad Krishna ;
Subramanian, Giridhar ;
Tan Hung Pham ;
Wang, Xiaofei ;
Perera, Shamira ;
Tun, Tin A. ;
Aung, Tin ;
Schmetterer, Leopold ;
Thiery, Alexandre H. ;
Girard, Michael J. A. .
SCIENTIFIC REPORTS, 2019, 9 (1)
[6]   Spectrally encoded coherence tomography and reflectometry: Simultaneous en face and cross-sectional imaging at 2 gigapixels per second [J].
El-Haddad, Mohamed T. ;
Bozic, Ivan ;
Tao, Yuankai K. .
JOURNAL OF BIOPHOTONICS, 2018, 11 (04)
[7]  
Fleishman G.M, 2017, MICCAI BRAINLES WORK, P43
[8]  
Frangi AF, 1998, LECT NOTES COMPUT SC, V1496, P130, DOI 10.1007/BFb0056195
[9]   Optical Coherence Tomography Angiography [J].
Gao, Simon S. ;
Jia, Yali ;
Zhang, Miao ;
Su, Johnny P. ;
Liu, Gangjun ;
Hwang, Thomas S. ;
Bailey, Steven T. ;
Huang, David .
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE, 2016, 57 (09) :OCT27-OCT36
[10]  
Giarratano Y, 2019, ARXIV PREPRINT ARXIV