Measurement of retinal nerve fiber layer thickness with a deep learning algorithm in ischemic optic neuropathy and optic neuritis
Abstract This work aims at determining the ability of a deep learning (DL) algorithm to measure retinal nerve fiber layer (RNFL) thickness from optical coherence tomography (OCT) scans in anterior ischemic optic neuropathy (NAION) and demyelinating optic neuritis (ON). The training/validation datase...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2022-10-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-022-22135-x |
| _version_ | 1798029847207870464 |
|---|---|
| author | Ghazale Razaghi Ehsan Hedayati Marjaneh Hejazi Rahele Kafieh Melika Samadi Robert Ritch Prem S. Subramanian Masoud Aghsaei Fard |
| author_facet | Ghazale Razaghi Ehsan Hedayati Marjaneh Hejazi Rahele Kafieh Melika Samadi Robert Ritch Prem S. Subramanian Masoud Aghsaei Fard |
| author_sort | Ghazale Razaghi |
| collection | DOAJ |
| description | Abstract This work aims at determining the ability of a deep learning (DL) algorithm to measure retinal nerve fiber layer (RNFL) thickness from optical coherence tomography (OCT) scans in anterior ischemic optic neuropathy (NAION) and demyelinating optic neuritis (ON). The training/validation dataset included 750 RNFL OCT B-scans. Performance of our algorithm was evaluated on 194 OCT B-scans from 70 healthy eyes, 82 scans from 28 NAION eyes, and 84 scans of 29 ON eyes. Results were compared to manual segmentation as a ground-truth and to RNFL calculations from the built-in instrument software. The Dice coefficient for the test images was 0.87. The mean average RNFL thickness using our U-Net was not different from the manually segmented best estimate and OCT machine data in control and ON eyes. In NAION eyes, while the mean average RNFL thickness using our U-Net algorithm was not different from the manual segmented value, the OCT machine data were different from the manual segmented values. In NAION eyes, the MAE of the average RNFL thickness was 1.18 ± 0.69 μm and 6.65 ± 5.37 μm in the U-Net algorithm segmentation and the conventional OCT machine data, respectively (P = 0.0001). |
| first_indexed | 2024-04-11T19:31:56Z |
| format | Article |
| id | doaj.art-7c0bfb3e0b304b878d90b090edb9ac2c |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-11T19:31:56Z |
| publishDate | 2022-10-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-7c0bfb3e0b304b878d90b090edb9ac2c2022-12-22T04:06:58ZengNature PortfolioScientific Reports2045-23222022-10-011211910.1038/s41598-022-22135-xMeasurement of retinal nerve fiber layer thickness with a deep learning algorithm in ischemic optic neuropathy and optic neuritisGhazale Razaghi0Ehsan Hedayati1Marjaneh Hejazi2Rahele Kafieh3Melika Samadi4Robert Ritch5Prem S. Subramanian6Masoud Aghsaei Fard7Medical Image Research Center, School of Advanced Technologies in Medicine, Tehran University of Medical SciencesFarabi Eye Hospital, Tehran University of Medical SciencesMedical Image Research Center, School of Advanced Technologies in Medicine, Tehran University of Medical SciencesDepatment of Engineering, Durham UniversityFarabi Eye Hospital, Tehran University of Medical SciencesEinhorn Clinical Research Center, New York Eye and Ear Infirmary of Mount SinaiDepartments of Ophthalmology, Neurology, and Neurosurgery, School of Medicine, University of ColoradoFarabi Eye Hospital, Tehran University of Medical SciencesAbstract This work aims at determining the ability of a deep learning (DL) algorithm to measure retinal nerve fiber layer (RNFL) thickness from optical coherence tomography (OCT) scans in anterior ischemic optic neuropathy (NAION) and demyelinating optic neuritis (ON). The training/validation dataset included 750 RNFL OCT B-scans. Performance of our algorithm was evaluated on 194 OCT B-scans from 70 healthy eyes, 82 scans from 28 NAION eyes, and 84 scans of 29 ON eyes. Results were compared to manual segmentation as a ground-truth and to RNFL calculations from the built-in instrument software. The Dice coefficient for the test images was 0.87. The mean average RNFL thickness using our U-Net was not different from the manually segmented best estimate and OCT machine data in control and ON eyes. In NAION eyes, while the mean average RNFL thickness using our U-Net algorithm was not different from the manual segmented value, the OCT machine data were different from the manual segmented values. In NAION eyes, the MAE of the average RNFL thickness was 1.18 ± 0.69 μm and 6.65 ± 5.37 μm in the U-Net algorithm segmentation and the conventional OCT machine data, respectively (P = 0.0001).https://doi.org/10.1038/s41598-022-22135-x |
| spellingShingle | Ghazale Razaghi Ehsan Hedayati Marjaneh Hejazi Rahele Kafieh Melika Samadi Robert Ritch Prem S. Subramanian Masoud Aghsaei Fard Measurement of retinal nerve fiber layer thickness with a deep learning algorithm in ischemic optic neuropathy and optic neuritis Scientific Reports |
| title | Measurement of retinal nerve fiber layer thickness with a deep learning algorithm in ischemic optic neuropathy and optic neuritis |
| title_full | Measurement of retinal nerve fiber layer thickness with a deep learning algorithm in ischemic optic neuropathy and optic neuritis |
| title_fullStr | Measurement of retinal nerve fiber layer thickness with a deep learning algorithm in ischemic optic neuropathy and optic neuritis |
| title_full_unstemmed | Measurement of retinal nerve fiber layer thickness with a deep learning algorithm in ischemic optic neuropathy and optic neuritis |
| title_short | Measurement of retinal nerve fiber layer thickness with a deep learning algorithm in ischemic optic neuropathy and optic neuritis |
| title_sort | measurement of retinal nerve fiber layer thickness with a deep learning algorithm in ischemic optic neuropathy and optic neuritis |
| url | https://doi.org/10.1038/s41598-022-22135-x |
| work_keys_str_mv | AT ghazalerazaghi measurementofretinalnervefiberlayerthicknesswithadeeplearningalgorithminischemicopticneuropathyandopticneuritis AT ehsanhedayati measurementofretinalnervefiberlayerthicknesswithadeeplearningalgorithminischemicopticneuropathyandopticneuritis AT marjanehhejazi measurementofretinalnervefiberlayerthicknesswithadeeplearningalgorithminischemicopticneuropathyandopticneuritis AT rahelekafieh measurementofretinalnervefiberlayerthicknesswithadeeplearningalgorithminischemicopticneuropathyandopticneuritis AT melikasamadi measurementofretinalnervefiberlayerthicknesswithadeeplearningalgorithminischemicopticneuropathyandopticneuritis AT robertritch measurementofretinalnervefiberlayerthicknesswithadeeplearningalgorithminischemicopticneuropathyandopticneuritis AT premssubramanian measurementofretinalnervefiberlayerthicknesswithadeeplearningalgorithminischemicopticneuropathyandopticneuritis AT masoudaghsaeifard measurementofretinalnervefiberlayerthicknesswithadeeplearningalgorithminischemicopticneuropathyandopticneuritis |