GenVFNet: Generating Visual Field From Optical Coherence Tomography Angiography by Conditional Generative Adversarial Networks
Glaucoma investigations play an important role in glaucoma management. This study aimed to generate a visual field pattern deviation map (GenVF) from optical coherence tomography angiography (OCTA) using conditional generative adversarial networks (cGAN), namely GenVFNet. cGAN consisted of two compo...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10684695/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832595961074941952 |
|---|---|
| author | Anita Manassakorn Supatana Auethavekiat Vera Sa-Ing Sunee Chansangpetch Kitiya Ratanawongphaibul Nopphawan Uramphorn Visanee Tantisevi |
| author_facet | Anita Manassakorn Supatana Auethavekiat Vera Sa-Ing Sunee Chansangpetch Kitiya Ratanawongphaibul Nopphawan Uramphorn Visanee Tantisevi |
| author_sort | Anita Manassakorn |
| collection | DOAJ |
| description | Glaucoma investigations play an important role in glaucoma management. This study aimed to generate a visual field pattern deviation map (GenVF) from optical coherence tomography angiography (OCTA) using conditional generative adversarial networks (cGAN), namely GenVFNet. cGAN consisted of two components: generator and discriminator. The generator consisted of encoder and decoder blocks and the discriminator was the classification learner. The training data was Grad-CAM of OCTA (GradOCTA) of glaucomatous eyes. Only eyes with a visual field (VF) compatible with optic disc photograph (DP) and/or optical coherence tomography (OCT) were used. 100 eyes (90.9%) were used for training and the remaining 10 eyes (9.1%) for testing. To conform with the five severity levels in clinical diagnosis, we quantized the generated VF (GenRawVF) from GenVFNet into five levels and named the quantized image GenVF. In the experiment, GenRawVFs were compared with actual pattern deviation images (RealRawVF) using the structural similarity index measure (SSIM), normalized root mean square error (NRMSE), Fréchet inception distance (FID), and confusion matrix. The average <inline-formula> <tex-math notation="LaTeX">$\times 95$ </tex-math></inline-formula>%CI SSIM, NRMSE, and FID were <inline-formula> <tex-math notation="LaTeX">$0.87\times 0.02$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$0.10\times 0.6$ </tex-math></inline-formula>, and <inline-formula> <tex-math notation="LaTeX">$7.54\times 3.47$ </tex-math></inline-formula>, respectively. GenVFNet was able to construct GenVF, especially in the superior hemifield. In addition, GenVFNet was tested with VFs of eyes whose DP and/or OCT did not correlate with the available VF. According to the expert’s opinion, 81 eyes (65.8%) of GenVF were correlated with DP and/or OCT. This study demonstrated the potential of GenVFNet to construct a VF pattern deviation from the OCTA that would be a benefit for patients who cannot perform a reliable VF. |
| format | Article |
| id | doaj-art-56b9b5860029416aace51e31c5f2b684 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-56b9b5860029416aace51e31c5f2b6842025-01-18T00:01:32ZengIEEEIEEE Access2169-35362024-01-011214584514585610.1109/ACCESS.2024.346468210684695GenVFNet: Generating Visual Field From Optical Coherence Tomography Angiography by Conditional Generative Adversarial NetworksAnita Manassakorn0https://orcid.org/0000-0001-7483-8656Supatana Auethavekiat1https://orcid.org/0000-0002-6642-766XVera Sa-Ing2Sunee Chansangpetch3https://orcid.org/0000-0002-8996-2868Kitiya Ratanawongphaibul4Nopphawan Uramphorn5https://orcid.org/0000-0003-4077-9322Visanee Tantisevi6Center of Excellence in Glaucoma, Faculty of Medicine, Chulalongkorn University, Bangkok, ThailandDepartment of Electrical Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok, ThailandDepartment of Electrical and Computer Engineering, Faculty of Engineering, King Mongkut’s University of Technology North Bangkok, Bangkok, ThailandCenter of Excellence in Glaucoma, Faculty of Medicine, Chulalongkorn University, Bangkok, ThailandCenter of Excellence in Glaucoma, Faculty of Medicine, Chulalongkorn University, Bangkok, ThailandDepartment of Ophthalmology, Faculty of Medicine Vajira Hospital, Navamindradhiraj University, Bangkok, ThailandCenter of Excellence in Glaucoma, Faculty of Medicine, Chulalongkorn University, Bangkok, ThailandGlaucoma investigations play an important role in glaucoma management. This study aimed to generate a visual field pattern deviation map (GenVF) from optical coherence tomography angiography (OCTA) using conditional generative adversarial networks (cGAN), namely GenVFNet. cGAN consisted of two components: generator and discriminator. The generator consisted of encoder and decoder blocks and the discriminator was the classification learner. The training data was Grad-CAM of OCTA (GradOCTA) of glaucomatous eyes. Only eyes with a visual field (VF) compatible with optic disc photograph (DP) and/or optical coherence tomography (OCT) were used. 100 eyes (90.9%) were used for training and the remaining 10 eyes (9.1%) for testing. To conform with the five severity levels in clinical diagnosis, we quantized the generated VF (GenRawVF) from GenVFNet into five levels and named the quantized image GenVF. In the experiment, GenRawVFs were compared with actual pattern deviation images (RealRawVF) using the structural similarity index measure (SSIM), normalized root mean square error (NRMSE), Fréchet inception distance (FID), and confusion matrix. The average <inline-formula> <tex-math notation="LaTeX">$\times 95$ </tex-math></inline-formula>%CI SSIM, NRMSE, and FID were <inline-formula> <tex-math notation="LaTeX">$0.87\times 0.02$ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$0.10\times 0.6$ </tex-math></inline-formula>, and <inline-formula> <tex-math notation="LaTeX">$7.54\times 3.47$ </tex-math></inline-formula>, respectively. GenVFNet was able to construct GenVF, especially in the superior hemifield. In addition, GenVFNet was tested with VFs of eyes whose DP and/or OCT did not correlate with the available VF. According to the expert’s opinion, 81 eyes (65.8%) of GenVF were correlated with DP and/or OCT. This study demonstrated the potential of GenVFNet to construct a VF pattern deviation from the OCTA that would be a benefit for patients who cannot perform a reliable VF.https://ieeexplore.ieee.org/document/10684695/Artificial intelligenceconditional generative adversarial networkdeep learningglaucomaoptical coherence tomography angiographyvisual field |
| spellingShingle | Anita Manassakorn Supatana Auethavekiat Vera Sa-Ing Sunee Chansangpetch Kitiya Ratanawongphaibul Nopphawan Uramphorn Visanee Tantisevi GenVFNet: Generating Visual Field From Optical Coherence Tomography Angiography by Conditional Generative Adversarial Networks IEEE Access Artificial intelligence conditional generative adversarial network deep learning glaucoma optical coherence tomography angiography visual field |
| title | GenVFNet: Generating Visual Field From Optical Coherence Tomography Angiography by Conditional Generative Adversarial Networks |
| title_full | GenVFNet: Generating Visual Field From Optical Coherence Tomography Angiography by Conditional Generative Adversarial Networks |
| title_fullStr | GenVFNet: Generating Visual Field From Optical Coherence Tomography Angiography by Conditional Generative Adversarial Networks |
| title_full_unstemmed | GenVFNet: Generating Visual Field From Optical Coherence Tomography Angiography by Conditional Generative Adversarial Networks |
| title_short | GenVFNet: Generating Visual Field From Optical Coherence Tomography Angiography by Conditional Generative Adversarial Networks |
| title_sort | genvfnet generating visual field from optical coherence tomography angiography by conditional generative adversarial networks |
| topic | Artificial intelligence conditional generative adversarial network deep learning glaucoma optical coherence tomography angiography visual field |
| url | https://ieeexplore.ieee.org/document/10684695/ |
| work_keys_str_mv | AT anitamanassakorn genvfnetgeneratingvisualfieldfromopticalcoherencetomographyangiographybyconditionalgenerativeadversarialnetworks AT supatanaauethavekiat genvfnetgeneratingvisualfieldfromopticalcoherencetomographyangiographybyconditionalgenerativeadversarialnetworks AT verasaing genvfnetgeneratingvisualfieldfromopticalcoherencetomographyangiographybyconditionalgenerativeadversarialnetworks AT suneechansangpetch genvfnetgeneratingvisualfieldfromopticalcoherencetomographyangiographybyconditionalgenerativeadversarialnetworks AT kitiyaratanawongphaibul genvfnetgeneratingvisualfieldfromopticalcoherencetomographyangiographybyconditionalgenerativeadversarialnetworks AT nopphawanuramphorn genvfnetgeneratingvisualfieldfromopticalcoherencetomographyangiographybyconditionalgenerativeadversarialnetworks AT visaneetantisevi genvfnetgeneratingvisualfieldfromopticalcoherencetomographyangiographybyconditionalgenerativeadversarialnetworks |