Cystoid macular edema segmentation of Optical Coherence Tomography images using fully convolutional neural networks and fully connected CRFs

In this paper we present a new method for cystoid macular edema (CME) segmentation in retinal Optical Coherence Tomography (OCT) images, using a fully convolutional neural network (FCN) and a fully connected conditional random fields (dense CRFs). As a first step, the framework trains the FCN model to extract features from retinal layers in OCT images, which exhibit CME, and then segments CME regions using the trained model. Thereafter, dense CRFs are used to refine the segmentation according to the edema appearance. We have trained and tested the framework with OCT images from 10 patients with diabetic macular edema (DME). Our experimental results show that fluid and concrete macular edema areas were segmented with good adherence to boundaries. A segmentation accuracy of $0.61\pm 0.21$ (Dice coefficient) was achieved, with respect to the ground truth, which compares favourably with the previous state-of-the-art that used a kernel regression based method ($0.51\pm 0.34$). Our approach is versatile and we believe it can be easily adapted to detect other macular defects.

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

[2]  Mark W. Johnson Etiology and treatment of macular edema. , 2009, American journal of ophthalmology.

[3]  Sina Farsiu,et al.  Kernel regression based segmentation of optical coherence tomography images with diabetic macular edema. , 2015, Biomedical optics express.

[4]  J. Duker,et al.  Optical coherence tomography – current and future applications , 2013, Current Opinion in Ophthalmology.

[5]  Milan Sonka,et al.  Intra-retinal layer segmentation of 3D optical coherence tomography using coarse grained diffusion map , 2012, Medical Image Anal..

[6]  Trevor Darrell,et al.  Caffe: Convolutional Architecture for Fast Feature Embedding , 2014, ACM Multimedia.

[7]  James G Fujimoto,et al.  Choroidal thickness in normal eyes measured using Cirrus HD optical coherence tomography. , 2010, American journal of ophthalmology.

[8]  Vladlen Koltun,et al.  Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials , 2011, NIPS.

[9]  R. Klein,et al.  The Wisconsin Epidemiologic Study of Diabetic Retinopathy: XVII. The 14-year incidence and progression of diabetic retinopathy and associated risk factors in type 1 diabetes. , 1998, Ophthalmology.

[10]  Delia Cabrera Fernandez,et al.  Delineating fluid-filled region boundaries in optical coherence tomography images of the retina , 2005, IEEE Transactions on Medical Imaging.

[11]  Pratul P. Srinivasan,et al.  Fully automatic software for retinal thickness in eyes with diabetic macular edema from images acquired by cirrus and spectralis systems. , 2013, Investigative ophthalmology & visual science.

[12]  S. Kishi,et al.  Patterns of diabetic macular edema with optical coherence tomography. , 1999, American journal of ophthalmology.

[13]  E Reichel,et al.  Quantitative assessment of macular edema with optical coherence tomography. , 1995, Archives of ophthalmology.

[14]  BoykovYuri,et al.  Graph Cuts and Efficient N-D Image Segmentation , 2006 .

[15]  Joseph A. Izatt,et al.  Automatic segmentation of seven retinal layers in SDOCT images congruent with expert manual segmentation , 2010, Optics express.

[16]  B. Bouma,et al.  Improved signal-to-noise ratio in spectral-domain compared with time-domain optical coherence tomography. , 2003, Optics letters.

[17]  G. G. Stokes "J." , 1890, The New Yale Book of Quotations.

[18]  J. Fujimoto,et al.  Ultrahigh speed spectral / Fourier domain OCT ophthalmic imaging at 70,000 to 312,500 axial scans per second. , 2008, Optics express.

[19]  Pascal A. Dufour,et al.  Graph-Based Multi-Surface Segmentation of OCT Data Using Trained Hard and Soft Constraints , 2013, IEEE Transactions on Medical Imaging.

[20]  Vladimir Kolmogorov,et al.  What energy functions can be minimized via graph cuts? , 2002, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[21]  Ali Erginay,et al.  Diagnosis of macular pseudoholes and lamellar macular holes by optical coherence tomography. , 2004, American journal of ophthalmology.

[22]  T. Clark,et al.  DISCREPANCY BETWEEN FLUORESCEIN ANGIOGRAPHY AND OPTICAL COHERENCE TOMOGRAPHY IN DETECTION OF MACULAR DISEASE , 2008, Retina.

[23]  Radford M. Neal Pattern Recognition and Machine Learning , 2007, Technometrics.

[24]  T. Sørensen,et al.  A method of establishing group of equal amplitude in plant sociobiology based on similarity of species content and its application to analyses of the vegetation on Danish commons , 1948 .

[25]  Adrian Bradu,et al.  Master slave en-face OCT/SLO. , 2015, Biomedical optics express.

[26]  Sina Farsiu,et al.  Automatic segmentation of up to ten layer boundaries in SD-OCT images of the mouse retina with and without missing layers due to pathology. , 2014, Biomedical optics express.

[27]  Jitendra Malik,et al.  A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics , 2001, Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001.

[28]  Vascular Endothelial Growth Factor In Ocular Fluid of Patients with Diabetic Retinopathy and Other Retinal Disorders , 1995 .

[29]  Seyed-Ahmad Ahmadi,et al.  V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation , 2016, 2016 Fourth International Conference on 3D Vision (3DV).

[30]  D. R. Fulkerson,et al.  Maximal Flow Through a Network , 1956 .