Improving OCT Image Segmentation of Retinal Layers by Utilizing a Machine Learning Based Multistage System of Stacked Multiscale Encoders and Decoders
Optical coherence tomography (OCT)-based retinal imagery is often utilized to determine influential factors in patient progression and treatment, for which the retinal layers of the human eye are investigated to assess a patient’s health status and eyesight. In this contribution, we propose a machin...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-10-01
|
Series: | Bioengineering |
Subjects: | |
Online Access: | https://www.mdpi.com/2306-5354/10/10/1177 |
_version_ | 1797574684345106432 |
---|---|
author | Arunodhayan Sampath Kumar Tobias Schlosser Holger Langner Marc Ritter Danny Kowerko |
author_facet | Arunodhayan Sampath Kumar Tobias Schlosser Holger Langner Marc Ritter Danny Kowerko |
author_sort | Arunodhayan Sampath Kumar |
collection | DOAJ |
description | Optical coherence tomography (OCT)-based retinal imagery is often utilized to determine influential factors in patient progression and treatment, for which the retinal layers of the human eye are investigated to assess a patient’s health status and eyesight. In this contribution, we propose a machine learning (ML)-based multistage system of stacked multiscale encoders and decoders for the image segmentation of OCT imagery of the retinal layers to enable the following evaluation regarding the physiological and pathological states. Our proposed system’s results highlight its benefits compared to currently investigated approaches by combining commonly deployed methods from deep learning (DL) while utilizing deep neural networks (DNN). We conclude that by stacking multiple multiscale encoders and decoders, improved scores for the image segmentation task can be achieved. Our retinal-layer-based segmentation results in a final segmentation performance of up to <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>82.25</mn><mo>±</mo><mn>0.74</mn></mrow></semantics></math></inline-formula>% for the Sørensen–Dice coefficient, outperforming the current best single-stage model by <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>1.55</mn></mrow></semantics></math></inline-formula>% with a score of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>80.70</mn><mo>±</mo><mn>0.20</mn></mrow></semantics></math></inline-formula>%, given the evaluated peripapillary OCT data set. Additionally, we provide results on the data sets Duke SD-OCT, Heidelberg, and UMN to illustrate our model’s performance on especially noisy data sets. |
first_indexed | 2024-03-10T21:25:48Z |
format | Article |
id | doaj.art-bb3cb18cf8a74b59beb42de60f2e8c53 |
institution | Directory Open Access Journal |
issn | 2306-5354 |
language | English |
last_indexed | 2024-03-10T21:25:48Z |
publishDate | 2023-10-01 |
publisher | MDPI AG |
record_format | Article |
series | Bioengineering |
spelling | doaj.art-bb3cb18cf8a74b59beb42de60f2e8c532023-11-19T15:42:05ZengMDPI AGBioengineering2306-53542023-10-011010117710.3390/bioengineering10101177Improving OCT Image Segmentation of Retinal Layers by Utilizing a Machine Learning Based Multistage System of Stacked Multiscale Encoders and DecodersArunodhayan Sampath Kumar0Tobias Schlosser1Holger Langner2Marc Ritter3Danny Kowerko4Junior Professorship of Media Computing, Chemnitz University of Technology, 09107 Chemnitz, GermanyJunior Professorship of Media Computing, Chemnitz University of Technology, 09107 Chemnitz, GermanyProfessorship of Media Informatics, University of Applied Sciences Mittweida, 09648 Mittweida, GermanyProfessorship of Media Informatics, University of Applied Sciences Mittweida, 09648 Mittweida, GermanyJunior Professorship of Media Computing, Chemnitz University of Technology, 09107 Chemnitz, GermanyOptical coherence tomography (OCT)-based retinal imagery is often utilized to determine influential factors in patient progression and treatment, for which the retinal layers of the human eye are investigated to assess a patient’s health status and eyesight. In this contribution, we propose a machine learning (ML)-based multistage system of stacked multiscale encoders and decoders for the image segmentation of OCT imagery of the retinal layers to enable the following evaluation regarding the physiological and pathological states. Our proposed system’s results highlight its benefits compared to currently investigated approaches by combining commonly deployed methods from deep learning (DL) while utilizing deep neural networks (DNN). We conclude that by stacking multiple multiscale encoders and decoders, improved scores for the image segmentation task can be achieved. Our retinal-layer-based segmentation results in a final segmentation performance of up to <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>82.25</mn><mo>±</mo><mn>0.74</mn></mrow></semantics></math></inline-formula>% for the Sørensen–Dice coefficient, outperforming the current best single-stage model by <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>1.55</mn></mrow></semantics></math></inline-formula>% with a score of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>80.70</mn><mo>±</mo><mn>0.20</mn></mrow></semantics></math></inline-formula>%, given the evaluated peripapillary OCT data set. Additionally, we provide results on the data sets Duke SD-OCT, Heidelberg, and UMN to illustrate our model’s performance on especially noisy data sets.https://www.mdpi.com/2306-5354/10/10/1177ophthalmologyophthalmology diseasesOCT biomarkersOCT segmentationcomputer vision and pattern recognitionmachine learning |
spellingShingle | Arunodhayan Sampath Kumar Tobias Schlosser Holger Langner Marc Ritter Danny Kowerko Improving OCT Image Segmentation of Retinal Layers by Utilizing a Machine Learning Based Multistage System of Stacked Multiscale Encoders and Decoders Bioengineering ophthalmology ophthalmology diseases OCT biomarkers OCT segmentation computer vision and pattern recognition machine learning |
title | Improving OCT Image Segmentation of Retinal Layers by Utilizing a Machine Learning Based Multistage System of Stacked Multiscale Encoders and Decoders |
title_full | Improving OCT Image Segmentation of Retinal Layers by Utilizing a Machine Learning Based Multistage System of Stacked Multiscale Encoders and Decoders |
title_fullStr | Improving OCT Image Segmentation of Retinal Layers by Utilizing a Machine Learning Based Multistage System of Stacked Multiscale Encoders and Decoders |
title_full_unstemmed | Improving OCT Image Segmentation of Retinal Layers by Utilizing a Machine Learning Based Multistage System of Stacked Multiscale Encoders and Decoders |
title_short | Improving OCT Image Segmentation of Retinal Layers by Utilizing a Machine Learning Based Multistage System of Stacked Multiscale Encoders and Decoders |
title_sort | improving oct image segmentation of retinal layers by utilizing a machine learning based multistage system of stacked multiscale encoders and decoders |
topic | ophthalmology ophthalmology diseases OCT biomarkers OCT segmentation computer vision and pattern recognition machine learning |
url | https://www.mdpi.com/2306-5354/10/10/1177 |
work_keys_str_mv | AT arunodhayansampathkumar improvingoctimagesegmentationofretinallayersbyutilizingamachinelearningbasedmultistagesystemofstackedmultiscaleencodersanddecoders AT tobiasschlosser improvingoctimagesegmentationofretinallayersbyutilizingamachinelearningbasedmultistagesystemofstackedmultiscaleencodersanddecoders AT holgerlangner improvingoctimagesegmentationofretinallayersbyutilizingamachinelearningbasedmultistagesystemofstackedmultiscaleencodersanddecoders AT marcritter improvingoctimagesegmentationofretinallayersbyutilizingamachinelearningbasedmultistagesystemofstackedmultiscaleencodersanddecoders AT dannykowerko improvingoctimagesegmentationofretinallayersbyutilizingamachinelearningbasedmultistagesystemofstackedmultiscaleencodersanddecoders |