# Automatic Segmentation of the Retinal Nerve Fiber Layer by Means of Mathematical Morphology and Deformable Models in 2D Optical Coherence Tomography Imaging

^{1}

^{2}

^{3}

^{4}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Materials and Methods

- Preprocessing to reduce the speckle noise and undesired artifacts in peripapillary 2D B-scan OCT images. This filtering is mainly performed by morphological operators.
- Peak detection on positive and negative part of the vertical gradient. This column-wise process (1D scenario) provides a rough detection of the layer’s boundaries.
- Reduction of possible discontinuities in previous coarse detection by processing adjacent columns (2D scenario) with a sliding window low-pass filter.
- Refinement of previous result by the imposition of the physical characteristics of biological tissues. The properties of elasticity and rigidity of the solution are incorporated by using a deformable model (in a 2D scenario).

#### 2.1. Materials: OCT Database

#### 2.2. Preprocessing: Mathematical Morphology

#### 2.3. Boundary Segmentation: Active Contours

#### 2.4. RNFL Layer Segmentation Process

#### 2.5. Segmentation of Boundary #1-UB

#### 2.6. Segmentation of Boundary #2-AB

#### 2.7. Segmentation of Boundary #3-LB

#### 2.8. RNFL Thickness Calculation

## 3. Results

## 4. Discussion

#### 4.1. Robustness against Parameter Variations and Speckle Noise

#### 4.2. Analysis of the Performance of the Methods

## 5. Conclusions

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Conflicts of Interest

## References

- Weinreb, R.N.; Khaw, P.T. Primary open-angle glaucoma. Lancet
**2004**, 363, 1711–1720. [Google Scholar] [CrossRef] - Tham, Y.C.; Li, X.; Wong, T.Y.; Quigley, H.A.; Aung, T.; Cheng, C.Y. Global Prevalence of Glaucoma and Projections of Glaucoma Burden through 2040: A Systematic Review and Meta-Analysis. Ophthalmology
**2014**, 121, 2081–2090. [Google Scholar] [CrossRef] [PubMed] - Casson, R.J.; Chidlow, G.; Wood, J.P.; Crowston, J.G.; Goldberg, I. Definition of glaucoma: Clinical and experimental concepts. Clin. Exp. Ophthalmol.
**2012**, 40, 341–349. [Google Scholar] [CrossRef] [PubMed][Green Version] - Kwon, Y.H.; Fingert, J.H.; Kuehn, M.H.; Alward, W.L. Primary Open-Angle Glaucoma. N. Engl. J. Med.
**2009**, 360, 1113–1124. [Google Scholar] [CrossRef][Green Version] - Swanson, E.A.; Fujimoto, J.G. The ecosystem that powered the translation of OCT from fundamental research to clinical and commercial impact. Biomed. Opt. Express
**2017**, 8, 1638–1664. [Google Scholar] [CrossRef][Green Version] - Huang, D.; Swanson, E.; Lin, C.; Schuman, J.; Stinson, W.; Chang, W.; Hee, M.; Flotte, T.; Gregory, K.; Puliafito, C.; et al. Optical coherence tomography. Science
**1991**, 254, 1178–1181. [Google Scholar] [CrossRef][Green Version] - Kanngiesser, J.; Roth, B. Wavefront Shaping Concepts for Application in Optical Coherence Tomography—A Review. Sensors
**2020**, 20, 44. [Google Scholar] [CrossRef] - Badalà, F.; Nouri-Mahdavi, K.; Raoof, D.A.; Leeprechanon, N.; Law, S.K.; Caprioli, J. Optic disk and nerve fiber layer imaging to detect glaucoma. Am. J. Ophthalmol.
**2007**, 144, 724–732. [Google Scholar] [CrossRef][Green Version] - Alasil, T.; Wang, K.; Keane, P.A.; Lee, H.; Baniasadi, N.; de Boer, J.F.; Chen, T.C. Analysis of normal retinal nerve fiber layer thickness by age, sex, and race using spectral domain optical coherence tomography. J. Glaucoma
**2013**, 22, 532–541. [Google Scholar] [CrossRef] - Sommer, A.; Miller, N.R.; Pollack, I.; Maumenee, A.E.; George, T. The Nerve Fiber Layer in the Diagnosis of Glaucoma. Arch. Ophthalmol.
**1977**, 95, 2149–2156. [Google Scholar] [CrossRef] - Sommer, A.; Katz, J.; Quigley, H.A.; Miller, N.R.; Robin, A.L.; Richter, R.C.; Witt, K.A. Clinically Detectable Nerve Fiber Atrophy Precedes the Onset of Glaucomatous Field Loss. Arch. Ophthalmol.
**1991**, 109, 77–83. [Google Scholar] [CrossRef] [PubMed] - Duke SD-OCT. Available online: https://people.duke.edu/~sf59/RPEDC_Ophth_2013_dataset.htm (accessed on 29 September 2021).
- Chiu, S.J.; Allingham, M.J.; Mettu, P.S.; Cousins, S.W.; Izatt, J.A.; Farsiu, S. Kernel regression based segmentation of optical coherence tomography images with diabetic macular edema. Biomed. Opt. Express
**2015**, 6, 1172–1194. [Google Scholar] [CrossRef] [PubMed][Green Version] - MGU-Net. Available online: https://github.com/Jiaxuan-Li/MGU-Net (accessed on 29 September 2021).
- Li, J.; Jin, P.; Zhu, J.; Zou, H.; Xu, X.; Tang, M.; Zhou, M.; Gan, Y.; He, J.; Ling, Y.; et al. Multi-scale GCN-assisted two-stage network for joint segmentation of retinal layers and discs in peripapillary OCT images. Biomed. Opt. Express
**2021**, 12, 2204–2220. [Google Scholar] [CrossRef] [PubMed] - The Annotated Retinal OCT Images (AROI) Database. Available online: https://ipg.fer.hr/ipg/resources/oct_image_database (accessed on 29 September 2021).
- Melinščak, M.; Radmilović, M.; Vatavuk, Z.; Lončarić, S. Annotated retinal optical coherence tomography images (AROI) database for joint retinal layer and fluid segmentation. Automatika
**2021**, 62, 375–385. [Google Scholar] [CrossRef] - Optical Coherence Tomography Image Database (OCTID). Available online: https://dataverse.scholarsportal.info/dataverse/OCTID (accessed on 29 September 2021).
- Gholami, P.; Roy, P.; Parthasarathy, M.K.; Lakshminarayanan, V. OCTID: Optical Coherence Tomography Image Database. arXiv arXiv:cs.CV/1812.07056, 2019. [CrossRef]
- OCT Data & Color Fundus Images of Left & Right Eyes of 50 Healthy Persons. Available online: https://sites.google.com/site/hosseinrabbanikhorasgani/datasets-1/oct-fundus-right-left (accessed on 29 September 2021).
- Mahmudi, T.; Kafieh, R.; Rabbani, H.; Mehri, A.; Akhlaghi, M.R. Evaluation of asymmetry in right and left eyes of normal individuals using extracted features from optical coherence tomography and fundus images. J. Med. Signals Sens.
**2021**, 11, 12–23. [Google Scholar] [CrossRef] - Abràmoff, M.D.; Garvin, M.K.; Sonka, M. Retinal Imaging and Image Analysis. IEEE Rev. Biomed. Eng.
**2010**, 3, 169–208. [Google Scholar] [CrossRef] [PubMed][Green Version] - Quellec, G.; Lee, K.; Dolejsi, M.; Garvin, M.K.; Abramoff, M.D.; Sonka, M. Three-Dimensional Analysis of Retinal Layer Texture: Identification of Fluid-Filled Regions in SD-OCT of the Macula. IEEE Trans. Med. Imaging
**2010**, 29, 1321–1330. [Google Scholar] [CrossRef] [PubMed][Green Version] - Dufour, P.A.; Ceklic, L.; Abdillahi, H.; Schroder, S.; De Dzanet, S.; Wolf-Schnurrbusch, U.; Kowal, J. Graph-Based Multi-Surface Segmentation of OCT Data Using Trained Hard and Soft Constraints. IEEE Trans. Med. Imaging
**2013**, 32, 531–543. [Google Scholar] [CrossRef] - Beck, M.; Joshi, D.S.; Berger, L.; Klose, G.; De Zanet, S.; Mosinska, A.; Apostolopoulos, S.; Ebneter, A.; Zinkernagel, M.S.; Wolf, S.; et al. Comparison of Drusen Volume Assessed by Two Different OCT Devices. J. Clin. Med.
**2020**, 9, 2657. [Google Scholar] [CrossRef] - Xiang, D.; Tian, H.; Yang, X.; Shi, F.; Zhu, W.; Chen, H.; Chen, X. Automatic Segmentation of Retinal Layer in OCT Images with Choroidal Neovascularization. IEEE Trans. Image Process.
**2018**, 27, 5880–5891. [Google Scholar] [CrossRef] [PubMed] - Girish, G.; Thakur, B.; Chowdhury, S.R.; Kothari, A.R.; Rajan, J. Segmentation of intra-retinal cysts from optical coherence tomography images using a fully convolutional neural network model. IEEE J. Biomed. Health Inform.
**2019**, 23, 296–304. [Google Scholar] [CrossRef] [PubMed] - Xu, X.; Lee, K.; Zhang, L.; Sonka, M.; Abramoff, M.D. Stratified Sampling Voxel Classification for Segmentation of Intraretinal and Subretinal Fluid in Longitudinal Clinical OCT Data. IEEE Trans. Med. Imaging
**2015**, 34, 1616–1623. [Google Scholar] [CrossRef] - Bogunović, H.; Venhuizen, F.; Klimscha, S.; Apostolopoulos, S.; Bab-Hadiashar, A.; Bagci, U.; Beg, M.F.; Bekalo, L.; Chen, Q.; Ciller, C.; et al. RETOUCH: The retinal OCT fluid detection and segmentation benchmark and challenge. IEEE Trans. Med. Imaging
**2019**, 38, 1858–1874. [Google Scholar] [CrossRef] [PubMed] - Schmidt-Erfurth, U.; Klimscha, S.; Waldstein, S.; Bogunović, H. A view of the current and future role of optical coherence tomography in the management of age-related macular degeneration. Eye
**2017**, 31, 26–44. [Google Scholar] [CrossRef] [PubMed] - Cabrera Fernández, D. Delineating fluid-filled region boundaries in optical coherence tomography images of the retina. IEEE Trans. Med. Imaging
**2005**, 24, 929–945. [Google Scholar] [CrossRef] - Fuller, A.; Zawadzki, R.; Choi, S.; Wiley, D.; Werner, J.; Hamann, B. Segmentation of Three-dimensional Retinal Image Data. IEEE Trans. Vis. Comput. Graph.
**2007**, 13, 1719–1726. [Google Scholar] [CrossRef] - Baroni, M.; Fortunato, P.; La Torre, A. Towards quantitative analysis of retinal features in optical coherence tomography. Med. Eng. Phys.
**2007**, 29, 432–441. [Google Scholar] [CrossRef] - Boyer, K.; Herzog, A.; Roberts, C. Automatic recovery of the optic nervehead geometry in optical coherence tomography. IEEE Trans. Med. Imaging
**2006**, 25, 553–570. [Google Scholar] [CrossRef] [PubMed] - Wang, C.; Wang, Y.X.; Li, Y. Automatic Choroidal Layer Segmentation Using Markov Random Field and Level Set Method. IEEE J. Biomed. Health Inform.
**2017**, 21, 1694–1702. [Google Scholar] [CrossRef] [PubMed] - Yazdanpanah, A.; Hamarneh, G.; Smith, B.R.; Sarunic, M.V. Segmentation of Intra-Retinal Layers From Optical Coherence Tomography Images Using an Active Contour Approach. IEEE Trans. Med. Imaging
**2011**, 30, 484–496. [Google Scholar] [CrossRef] - Rossant, F.; Bloch, I.; Ghorbel, I.; Paques, M. Parallel Double Snakes. Application to the segmentation of retinal layers in 2D-OCT for pathological subjects. Pattern Recognit.
**2015**, 48, 3857–3870. [Google Scholar] [CrossRef] - Garvin, M.K.; Abramoff, M.D.; Wu, X.; Russell, S.R.; Burns, T.L.; Sonka, M. Automated 3-D Intraretinal Layer Segmentation of Macular Spectral-Domain Optical Coherence Tomography Images. IEEE Trans. Med. Imaging
**2009**, 28, 1436–1447. [Google Scholar] [CrossRef][Green Version] - Kafieh, R.; Rabbani, H.; Abramoff, M.D.; Sonka, M. Intra-retinal layer segmentation of 3D optical coherence tomography using coarse grained diffusion map. Med. Image Anal.
**2013**, 17, 907–928. [Google Scholar] [CrossRef][Green Version] - Novosel, J.; Thepass, G.; Lemij, H.G.; de Boer, J.F.; Vermeer, K.A.; van Vliet, L.J. Loosely coupled level sets for simultaneous 3D retinal layer segmentation in optical coherence tomography. Med. Image Anal.
**2015**, 26, 146–158. [Google Scholar] [CrossRef] - Dodo, B.I.; Li, Y.; Eltayef, K.; Liu, X. Automatic annotation of retinal layers in optical coherence tomography images. J. Med. Syst.
**2019**, 43, 1–10. [Google Scholar] [CrossRef] [PubMed][Green Version] - Duan, J.; Tench, C.; Gottlob, I.; Proudlock, F.; Bai, L. Automated segmentation of retinal layers from optical coherence tomography images using geodesic distance. Pattern Recognit.
**2017**, 72, 158–175. [Google Scholar] [CrossRef] - Rong, Y.; Xiang, D.; Zhu, W.; Yu, K.; Shi, F.; Fan, Z.; Chen, X. Surrogate-assisted retinal OCT image classification based on convolutional neural networks. IEEE J. Biomed. Health Inform.
**2019**, 23, 253–263. [Google Scholar] [CrossRef] - García, G.; del Amor, R.; Colomer, A.; Verdú-Monedero, R.; Morales-Sánchez, J.; Naranjo, V. Circumpapillary OCT-focused hybrid learning for glaucoma grading using tailored prototypical neural networks. Artif. Intell. Med.
**2021**, 118, 102132. [Google Scholar] [CrossRef] - Fang, L.; Cunefare, D.; Wang, C.; Guymer, R.H.; Li, S.; Farsiu, S. Automatic segmentation of nine retinal layer boundaries in OCT images of non-exudative AMD patients using deep learning and graph search. Biomed. Opt. Express
**2017**, 8, 2732–2744. [Google Scholar] [CrossRef] [PubMed][Green Version] - Li, Q.; Li, S.; He, Z.; Guan, H.; Chen, R.; Xu, Y.; Wang, T.; Qi, S.; Mei, J.; Wang, W. DeepRetina: Layer Segmentation of Retina in OCT Images Using Deep Learning. Transl. Vis. Sci. Technol.
**2020**, 9, 1–17. [Google Scholar] [CrossRef] - Xiang, D.; Chen, G.; Shi, F.; Zhu, W.; Liu, Q.; Yuan, S.; Chen, X. Automatic Retinal Layer Segmentation of OCT Images With Central Serous Retinopathy. IEEE J. Biomed. Health Inform.
**2019**, 23, 283–295. [Google Scholar] [CrossRef] - Morales, S.; Colomer, A.; Mossi, J.M.; Del Amor, R.; Woldbye, D.; Klemp, K.; Larsen, M.; Naranjo, V. Retinal layer segmentation in rodent OCT images: Local intensity profiles & fully convolutional neural networks. Comput. Methods Programs Biomed.
**2021**, 198, 105788. [Google Scholar] [CrossRef] - Liu, X.; Cao, J.; Wang, S.; Zhang, Y.; Wang, M. Confidence-Guided Topology-Preserving Layer Segmentation for Optical Coherence Tomography Images With Focus-Column Module. IEEE Trans. Instrum. Meas.
**2021**, 70, 1–12. [Google Scholar] [CrossRef] - Soille, P. Morphological Image Analysis: Principles and Applications, 2nd ed.; Springer: Berlin/Heidelberg, Germany, 2003. [Google Scholar]
- Weruaga, L.; Verdú, R.; Morales, J. Frequency Domain Formulation of Active Parametric Deformable Models. IEEE Trans. Pattern Anal. Mach. Intell.
**2004**, 26, 1568–1578. [Google Scholar] [CrossRef] [PubMed] - Berenguer-Vidal, R.; Verdú-Monedero, R.; Morales-Sánchez, J. Design of B-spline multidimensional deformable models in the frequency domain. Math. Comput. Model.
**2012**, 57, 1942–1949. [Google Scholar] [CrossRef] - Verdú-Monedero, R.; Morales-Sánchez, J.; Weruaga, L. Convergence analysis of active contours. Image Vis. Comput.
**2008**, 26, 1118–1128. [Google Scholar] [CrossRef] - Berenguer-Vidal, R.; Verdú-Monedero, R.; Morales-Sánchez, J. Convergence analysis of multidimensional parametric deformable models. Comput. Vis. Image Underst.
**2015**, 135, 157–177. [Google Scholar] [CrossRef] - Serra, J. Image Analysis and Mathematical Morphology; Academic Press: London, UK, 1982; Volume I. [Google Scholar]
- Serra, J. Image Analysis and Mathematical Morphology: Theoretical Advances; Academic Press: London, UK, 1988; Volume II. [Google Scholar]
- Xu, C.; Prince, J.L. Gradient Vector Flow Deformable Models; Handbook of Medical Imaging, Isaac Bankman; Academic Press: Cambridge, MA, USA, 2000. [Google Scholar]
- Terzopoulos, D. Deformable models: Classic, topology-adaptive and generalized formulations. In Geometric Level Set Methods in Imaging, Vision and Graphics; Osher, S., Paragios, N., Eds.; Springer: New York, NY, USA, 2003; Chapter 2; pp. 21–40. [Google Scholar]
- Liang, J.; McInerney, T.; Terzopoulos, D. United snakes. Med. Image Anal.
**2006**, 10, 215–233. [Google Scholar] [CrossRef] - Bastida-Jumilla, M.; Menchon-Lara, R.; Morales-Sanchez, J.; Verdu-Monedero, R.; Larrey-Ruiz, J.; Sancho-Gomez, J. Frequency-domain active contours solution to evaluate intima-media thickness of the common carotid artery. Biomed. Signal Process. Control
**2015**, 16, 68–79. [Google Scholar] [CrossRef] - Kass, M.; Witkin, A.; Terzopoulos, D. Snakes: Active contour models. Int. J. Comput. Vis.
**1988**, 1, 321–331. [Google Scholar] [CrossRef] - Terzopoulos, D. The computation of visible surface representations. IEEE Trans. Pattern Anal. Mach. Intell.
**1988**, 10, 417–438. [Google Scholar] [CrossRef] - Malladi, R.; Sethian, J.; Vemuri, B. Shape modeling with front propagation: A level set approach. IEEE Trans. Pattern Anal. Mach. Intell.
**1995**, 17, 158–175. [Google Scholar] [CrossRef][Green Version] - Cremers, D.; Schnorr, C.; Weickert, J. Diffusion-snakes: Combining statistical shape knowledge and image information in a variational framework. In Proceedings of the 1st IEEE Workshop on Variational and Level Set Methods in Computer Vision, Vancouver, BC, Canada, 13 July 2001; pp. 137–144. [Google Scholar]
- Wei, P.; Wang, W.; Yang, Y.; Wang, M.Y. Level set band method: A combination of density-based and level set methods for the topology optimization of continuums. Front. Mech. Eng.
**2020**, 15, 390–405. [Google Scholar] [CrossRef] - Unser, M.; Aldroubi, A.; Eden, M. B-spline signal processing: Part I - Theory; Part II - Efficient design and applications. IEEE Trans. Signal Process.
**1993**, 41, 821–848. [Google Scholar] [CrossRef] - Oppenheim, A.; Schafer, R. Discrete-Time Signal Processing, 2nd ed.; Prentice-Hall: Upper Saddle River, NJ, USA, 1999. [Google Scholar]
- Gonzalez, R.C.; Woods, R.E. Digital Image Processing, 3rd ed.; Prentice-Hall, Inc.: Upper Saddle River, NJ, USA, 2006. [Google Scholar]
- Amor, R.d.; Morales, S.; Colomer, A.n.; Mossi, J.M.; Woldbye, D.; Klemp, K.; Larsen, M.; Naranjo, V. Towards Automatic Glaucoma Assessment: An Encoder-decoder CNN for Retinal Layer Segmentation in Rodent OCT images. In Proceedings of the 27th European Signal Processing Conference (EUSIPCO), A Coruña, Spain, 2–6 September 2019; pp. 1–5. [Google Scholar] [CrossRef][Green Version]
- Dice, L.R. Measures of the amount of ecologic association between species. Ecology
**1945**, 26, 297–302. [Google Scholar] [CrossRef] - Laligant, O.; Truchetet, F.; Fauvet, E. Noise estimation from digital step-model signal. IEEE Trans. Image Process.
**2013**, 22, 5158–5167. [Google Scholar] [CrossRef]

**Figure 1.**Peripapillary B-scan OCT centered on the optic disc and its relative axial position with respect to a retinal fundus image. Screenshot taken from the Spectralis OCT device. I, S, N, and T stand for inferior, superior, nasal, and temporal, respectively.

**Figure 2.**Screenshot provided by the Spectralis software version 6.9.4.0. From left to right and top to bottom: Retinal fundus photography centered on the optic disc (the yellow circle indicates the location of the peripapillary B-scan, which is shown on the right with the segmentation of the RNFL); 2D peripapillary B-scan OCT on Cartesian coordinates; estimated mean values for RNFL layer thickness for the temporal (T), temporal superior (TS), nasal superior (NS), nasal (N), nasal inferior (NI), and temporal inferior (TI) sectors, as well as the overall mean (G); rectified outline of the RNFL with estimated thickness and reference values according to the database European Descent (2009).

**Figure 3.**Example of a peripapillary B-scan OCT image with the manual segmentation of the layer of interest RNFL and description of the sectors: T, TS, NS, N, NI, and TI.

**Figure 4.**Layers of the retina from top to bottom. RNFL: retinal nerve fiber layer; GCL: ganglion cell layer; IPL: inner plexiform layer; INL: inner nuclear layer; OPL: outer plexiform layer; and ONL: outer nuclear layer. Boundary delineation in the segmentation process. #1-UB: upper boundary of the RNFL; #2-AB: auxiliary boundary corresponding to the lower boundary of the ONL; #3-LB, lower boundary of the RNFL. The dashed boxes exemplify some of the artifacts of the layers.

**Figure 5.**Retinal fundus imaging of the eye under analysis. The green circumference represents the OCT analysis circumference centered on the optic nerve, marked with a

**+**. The numbers represent the vessels crossing this analysis circumference.

**Figure 6.**Intensity inhomogeneities and boundary artifacts caused by shadows cast by blood vessels. The numbers identify each of the vessels crossing the circular ray tracing depicted in Figure 5.

**Figure 9.**Selection of the #2-AB vertical position. On the left, the positive vertical gradient image ${I}_{9}^{\#2}$. On the right, 1D profile of the dashed column of ${I}_{9}^{\#2}$.

**Figure 11.**Representation of the filter used in the first stage of blood vessel suppression: (

**a**) space domain representation of the filter or point spread function (PSF), and (

**b**) frequency domain representation or transfer function.

**Figure 13.**Selection of the #3-LB vertical position. On the left, the negative vertical gradient image ${I}_{9}^{\#3}$. On the right, 1D profile of the dashed column of ${I}_{9}^{\#3}$.

**Figure 14.**RNFL thickness measurements for each sector of the peripapillary circumference. On the left, diagram detailing the value of the average thickness of the RNFL in each sector of the eye. On the right side, representation of the borders #1-UB and #3-LB, and the average RNFL thickness in these sectors.

**Figure 15.**Segmentation results for the right eye of (

**a**–

**d**) patient 2, (

**e**–

**h**) patient 20, and (

**i**–

**l**) patient 69.

**Figure 16.**Results of changing the radius of the structuring element from ${\u25ef}_{7}$ to ${\u25ef}_{5}$ and ${\u25ef}_{9}$, used in the steps illustrated in Figure 7b–d for the delineation of #1-UB.

**Figure 17.**Results of changing the radius of the structuring element from ${\u25ef}_{3}$ to ${\u25ef}_{1}$ and ${\u25ef}_{7}$, used in the steps illustrated in Figure 8h–k for the delineation of #2-AB.

**Figure 18.**Results of testing the robustness against synthetically incorporated speckle noise from ${\sigma}^{2}=0.001$ to ${\sigma}^{2}=0.1$ ($\sigma =0.031$ to $\sigma =0.316$). The NOLSE estimation [71] of the speckle noise for the raw image is ${\sigma}^{2}=0.0000381$ ($\sigma =0.0062$).

Eye | Healthy | Suspicious | Unhealthy |
---|---|---|---|

Left | 88 | 31 | 43 |

Right | 97 | 32 | 38 |

Sector | Polar (Degrees) | Cartesian (Pixels) | ||
---|---|---|---|---|

Min | Max | Min | Max | |

Temporal (T) | 0° | 45° | 1 | 96 |

Temporal Superior (TS) | 45° | 90° | 97 | 192 |

Nasal Superior (NS) | 90° | 135° | 193 | 288 |

Nasal (N) | 135° | 225° | 289 | 480 |

Nasal Inferior (NI) | 225° | 270° | 481 | 576 |

Temporal Inferior (TI) | 270° | 315° | 577 | 672 |

Temporal (T) | 315° | 360° | 673 | 768 |

**Table 3.**Dice similarity coefficient calculated comparing the manual segmentation of the expert with results of the proposed method and H-DLpNet method for each sector of the eye and for the whole layer, G, for the right eye of patients 2, 20, and 69.

Dice Similarity Coefficient | ||||||
---|---|---|---|---|---|---|

Patient 2 | Patient 20 | Patient 69 | ||||

Sectors | Proposed | H-DLpNet | Proposed | H-DLpNet | Proposed | H-DLpNet |

Method | Method | Method | ||||

TS | $0.974$ | $0.951$ | $0.943$ | $0.927$ | $0.899$ | $0.832$ |

T | $0.959$ | $0.927$ | $0.929$ | $0.857$ | $0.927$ | $0.887$ |

TI | $0.955$ | $0.954$ | $0.941$ | $0.926$ | $0.947$ | $0.950$ |

NS | $0.959$ | $0.912$ | $0.818$ | $0.873$ | $0.851$ | $0.820$ |

N | $0.934$ | $0.902$ | $0.896$ | $0.893$ | $0.895$ | $0.850$ |

NI | $0.965$ | $0.914$ | $0.943$ | $0.926$ | $0.924$ | $0.886$ |

G | $0.956$ | $0.927$ | $0.916$ | $0.901$ | $0.908$ | $0.874$ |

**Table 4.**RNFL thickness (mean and standard deviation in µm) provided by the proposed method, the H-DLpNet method and Spectralis for each sector of the eye (TS, T, TI, NS, N, and NI) as well as for the whole layer (G).

Sectors | RNFL Thickness Measurement (µm) | |||
---|---|---|---|---|

Proposed | H-DLpNet | Spectralis | ||

Left eye | TS | 122.8 ± 23.6 | 128.5 ± 27.4 | 121.7 ± 30.5 |

T | 79.0 ± 11.6 | 70.7 ± 13.9 | 67.3 ± 13.3 | |

TI | 136.2 ± 30.2 | 132.7 ± 31.8 | 127.4 ± 34.8 | |

NS | 116.0 ± 21.4 | 111.4 ± 28.1 | 105.5 ± 30.2 | |

N | 86.5 ± 15.7 | 74.7 ± 16.4 | 68.7 ± 19.1 | |

NI | 114.2 ± 23.6 | 109.1 ± 27.1 | 102.2 ± 28.0 | |

G | 102.5 ± 13.2 | 96.6 ± 17.0 | 90.9 ± 17.8 | |

Right eye | TS | 126.5 ± 24.9 | 129.8 ± 25.5 | 125.2 ± 29.5 |

T | 80.9 ± 14.0 | 72.7 ± 15.6 | 68.6 ± 15.6 | |

TI | 133.4 ± 26.0 | 131.9 ± 29.9 | 126.0 ± 34.3 | |

NS | 112.0 ± 23.5 | 106.0 ± 27.6 | 98.6 ± 27.4 | |

N | 90.9 ± 14.9 | 79.6 ± 16.3 | 72.8 ± 17.9 | |

NI | 113.4 ± 23.5 | 105.9 ± 25.8 | 102.6 ± 27.3 | |

G | 103.6 ± 13.6 | 97.3 ± 16.7 | 92.0 ± 17.8 |

**Table 5.**Absolute thickness error (mean and standard deviation in µm) calculated as the difference of the RNFL thickness provided by the proposed approach to both H-DLpNet and Spectralis methods. Relative error (mean and standard deviation) computed for each sector as the absolute error divided by the thickness value determined by the proposed method. These values are calculated for each sector of the eye (TS, T, TI, NS, N, and NI) as well as for the whole layer (G).

Sectors | Absolute Thickness Errors (µm) | Relative Thickness Errors | |||
---|---|---|---|---|---|

H-DLpNet | Spectralis | H-DLpNet | Spectralis | ||

Left eye | TS | 13.1 ± 11.8 | 14.5 ± 12.9 | 0.12 ± 0.11 | 0.13 ± 0.12 |

T | 8.3 ± 6.1 | 11.8 ± 7.8 | 0.11 ± 0.09 | 0.15 ± 0.10 | |

TI | 10.0 ± 18.1 | 11.9 ± 10.0 | 0.08 ± 0.08 | 0.10 ± 0.11 | |

NS | 12.3 ± 10.9 | 14.9 ± 12.6 | 0.11 ± 0.10 | 0.14 ± 0.12 | |

N | 11.7 ± 11.0 | 17.6 ± 13.3 | 0.13 ± 0.12 | 0.20 ± 0.14 | |

NI | 10.3 ± 9.2 | 13.5 ± 11.4 | 0.10 ± 0.09 | 0.13 ± 0.11 | |

G | 6.9 ± 6.1 | 10.8 ± 8.0 | 0.07 ± 0.07 | 0.11 ± 0.09 | |

Right eye | TS | 12.1 ± 10.4 | 12.8 ± 11.2 | 0.10 ± 0.10 | 0.11 ± 0.10 |

T | 8.3 ± 6.0 | 12.1 ± 7.9 | 0.11 ± 0.08 | 0.15 ± 0.10 | |

TI | 12.8 ± 11.0 | 15.2 ± 12.4 | 0.11 ± 0.10 | 0.13 ± 0.12 | |

NS | 10.1 ± 9.5 | 15.2 ± 13.5 | 0.10 ± 0.09 | 0.14 ± 0.12 | |

N | 11.3 ± 9.3 | 17.3 ± 12.9 | 0.13 ± 0.10 | 0.19 ± 0.14 | |

NI | 10.9 ± 9.6 | 13.0 ± 10.5 | 0.10 ± 0.09 | 0.12 ± 0.10 | |

G | 7.4 ± 6.4 | 10.4 ± 7.3 | 0.08 ± 0.07 | 0.11 ± 0.08 |

**Table 6.**Dice similarity coefficient (mean and standard deviation) calculated comparing the results of the proposed method and the H-DLpNet method for each sector of the eye and for the whole layer.

Sectors | Dice Similarity Coefficient | |
---|---|---|

H-DLpNet | ||

Left eye | TS | 0.895 ± 0.082 |

T | 0.918 ± 0.071 | |

TI | 0.921 ± 0.057 | |

NS | 0.898 ± 0.059 | |

N | 0.893 ± 0.078 | |

NI | 0.917 ± 0.057 | |

G | 0.908 ± 0.047 | |

Right eye | TS | 0.906 ± 0.059 |

T | 0.919 ± 0.053 | |

TI | 0.914 ± 0.074 | |

NS | 0.907 ± 0.054 | |

N | 0.902 ± 0.063 | |

NI | 0.917 ± 0.062 | |

G | 0.911 ± 0.047 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Berenguer-Vidal, R.; Verdú-Monedero, R.; Morales-Sánchez, J.; Sellés-Navarro, I.; del Amor, R.; García, G.; Naranjo, V.
Automatic Segmentation of the Retinal Nerve Fiber Layer by Means of Mathematical Morphology and Deformable Models in 2D Optical Coherence Tomography Imaging. *Sensors* **2021**, *21*, 8027.
https://doi.org/10.3390/s21238027

**AMA Style**

Berenguer-Vidal R, Verdú-Monedero R, Morales-Sánchez J, Sellés-Navarro I, del Amor R, García G, Naranjo V.
Automatic Segmentation of the Retinal Nerve Fiber Layer by Means of Mathematical Morphology and Deformable Models in 2D Optical Coherence Tomography Imaging. *Sensors*. 2021; 21(23):8027.
https://doi.org/10.3390/s21238027

**Chicago/Turabian Style**

Berenguer-Vidal, Rafael, Rafael Verdú-Monedero, Juan Morales-Sánchez, Inmaculada Sellés-Navarro, Rocío del Amor, Gabriel García, and Valery Naranjo.
2021. "Automatic Segmentation of the Retinal Nerve Fiber Layer by Means of Mathematical Morphology and Deformable Models in 2D Optical Coherence Tomography Imaging" *Sensors* 21, no. 23: 8027.
https://doi.org/10.3390/s21238027