{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,23]],"date-time":"2026-06-23T23:47:51Z","timestamp":1782258471450,"version":"3.54.5"},"reference-count":37,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,2,3]],"date-time":"2025-02-03T00:00:00Z","timestamp":1738540800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Ophthalmol."],"abstract":"<jats:sec><jats:title>Introduction<\/jats:title><jats:p>Glaucoma, optic neuritis (ON), and non-arteritic anterior ischemic optic neuropathy (NAION) produce distinct patterns of retinal ganglion cell (RGC) damage. We propose a booster Variational Autoencoder (bVAE) to capture spatial variations in RGC loss and generate latent space (LS) montage maps that visualize different degrees and spatial patterns of optic nerve bundle injury. Furthermore, the bVAE model is capable of tracking the spatial pattern of RGC thinning over time and classifying the underlying cause.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>The bVAE model consists of an encoder, a display decoder, and a booster decoder. The encoder decomposes input ganglion cell layer (GCL) thickness maps into two display latent variables (dLVs) and eight booster latent variables (bLVs). The dLVs capture primary spatial patterns of RGC thinning, while the display decoder reconstructs the GCL map and creates the LS montage map. The bLVs add finer spatial details, improving reconstruction accuracy. XGBoost was used to analyze the dLVs and bLVs, estimating normal\/abnormal GCL thinning and classifying diseases (glaucoma, ON, and NAION). A total of 10,701 OCT macular scans from 822 subjects were included in this study.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Incorporating bLVs improved reconstruction accuracy, with the image-based root-mean-square error (RMSE) between input and reconstructed GCL thickness maps decreasing from 5.55 \u00b1 2.29 \u00b5m (two dLVs only) to 4.02 \u00b1 1.61 \u00b5m (two dLVs and eight bLVs). However, the image-based structural similarity index (SSIM) remained similar (0.91 \u00b1 0.04), indicating that just two dLVs effectively capture the main GCL spatial patterns. For classification, the XGBoost model achieved an AUC of 0.98 for identifying abnormal spatial patterns of GCL thinning over time using the dLVs. Disease classification yielded AUCs of 0.95 for glaucoma, 0.84 for ON, and 0.93 for NAION, with bLVs further increasing the AUCs to 0.96 for glaucoma, 0.93 for ON, and 0.99 for NAION.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusion<\/jats:title><jats:p>This study presents a novel approach to visualizing and quantifying GCL thinning patterns in optic neuropathies using the bVAE model. The combination of dLVs and bLVs enhances the model\u2019s ability to capture key spatial features and predict disease progression. Future work will focus on integrating additional image modalities to further refine the model\u2019s diagnostic capabilities.<\/jats:p><\/jats:sec>","DOI":"10.3389\/fopht.2024.1497848","type":"journal-article","created":{"date-parts":[[2025,2,3]],"date-time":"2025-02-03T06:36:13Z","timestamp":1738564573000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":6,"title":["Quantifying the spatial patterns of retinal ganglion cell loss and progression in optic neuropathy by applying a deep learning variational autoencoder approach to optical coherence tomography"],"prefix":"10.3389","volume":"4","author":[{"given":"Jui-Kai","family":"Wang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Brett A.","family":"Johnson","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhi","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Honghai","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"David","family":"Szanto","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Brian","family":"Woods","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael","family":"Wall","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Young H.","family":"Kwon","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Edward F.","family":"Linton","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Andrew","family":"Pouw","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mark J.","family":"Kupersmith","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mona K.","family":"Garvin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Randy H.","family":"Kardon","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,2,3]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","DOI":"10.1097\/WNO.0b013e318238b9cb","article-title":"Role of the macular optical coherence tomography scan in neuro-ophthalmology","volume":"31","author":"Kardon","year":"2011","journal-title":"J Neuro-Ophthalmology"},{"key":"B2","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.preteyeres.2012.08.003","article-title":"Glaucomatous damage of the macula","volume":"32","author":"Hood","year":"2013","journal-title":"Prog Retin Eye Res"},{"key":"B3","doi-asserted-by":"publisher","DOI":"10.1167\/iovs.15-18736","article-title":"Retinal ganglion cell layer thinning within one month of presentation for non-arteritic anterior ischemic optic neuropathy","volume":"57","author":"Kupersmith","year":"2016","journal-title":"Invest Ophthalmol Visual Sci"},{"key":"B4","doi-asserted-by":"publisher","DOI":"10.1177\/1352458515598020","article-title":"Retinal ganglion cell layer thinning within one month of presentation for optic neuritis","volume":"22","author":"Kupersmith","year":"2016","journal-title":"Mult Scler"},{"key":"B5","doi-asserted-by":"publisher","first-page":"35","DOI":"10.21037\/aes.2018.05.08","article-title":"The role of optical coherence tomography in neuro-ophthalmology","volume":"3","author":"Chen","year":"2018","journal-title":"Ann Eye Sci"},{"key":"B6","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-030-87000-3_17","article-title":"Representation and reconstruction of image-based structural patterns of glaucomatous defects using only two latent variables from a variational autoencoder","volume-title":"Ophthalmic medical image analysis (OMIA)","author":"Wang","year":"2021"},{"key":"B7","doi-asserted-by":"publisher","DOI":"10.1097\/IJG.0000000000002348","article-title":"Comparison of patterns of structural progression in primary open angle glaucoma and pseudoexfoliation glaucoma","volume":"33","author":"Wy","year":"2024","journal-title":"J Glaucoma"},{"key":"B8","first-page":"1","article-title":"Ms minute: Retinal optical coherence tomography for ms","volume":"30","author":"Schorr","year":"2022","journal-title":"Pratical Neurol"},{"key":"B9","doi-asserted-by":"publisher","DOI":"10.2147\/OPTH.S42522","article-title":"Long-term retinal nerve fiber layer changes following nonarteritic anterior ischemic optic neuropathy","volume":"7","author":"Dotan","year":"2013","journal-title":"Clin Ophthalmol"},{"key":"B10","doi-asserted-by":"publisher","DOI":"10.1364\/BOE.516045","article-title":"Hybrid deep learning and optimal graph search method for optical coherence tomography layer segmentation in diseases affecting the optic nerve","volume":"15","author":"Chen","year":"2024","journal-title":"Biomed Optics Express"},{"key":"B11","doi-asserted-by":"publisher","first-page":"9541","DOI":"10.1038\/s41598-020-66355-5","article-title":"Automated retinal layer segmentation using graph-based algorithm incorporating deep-learning-derived information","volume":"10","author":"Mishra","year":"2020","journal-title":"Sci Rep"},{"key":"B12","doi-asserted-by":"publisher","DOI":"10.3390\/jimaging8050139","article-title":"Intraretinal layer segmentation using cascaded compressed u-nets","volume":"8","author":"Yadav","year":"2022","journal-title":"J Imaging"},{"key":"B13","doi-asserted-by":"publisher","DOI":"10.1364\/BOE.487518","article-title":"Longitudinal deep network for consistent OCT layer segmentation","volume":"14","author":"He","year":"2023","journal-title":"BioMed Opt Express"},{"key":"B14","doi-asserted-by":"publisher","DOI":"10.1167\/iovs.11-7962","article-title":"Macular ganglion cell\u2013inner plexiform layer: Automated detection and thickness reproducibility with spectral domain\u2013optical coherence tomography in glaucoma","volume":"52","author":"Mwanza","year":"2011","journal-title":"Invest Ophthalmol Visual Sci"},{"key":"B15","doi-asserted-by":"publisher","DOI":"10.1561\/2200000056","article-title":"An introduction to variational autoencoders","volume":"12","author":"Kingma","year":"2019","journal-title":"Foundations Trends Mach Learn"},{"key":"B16","first-page":"1","article-title":"Understanding disentangling in \u03b2-VAE","volume-title":"Neural information processing systems (NeurIPS) workshop on learning disentangled representations","author":"Burgess","year":"2018"},{"key":"B17","article-title":"Deep feature consistent variational autoencoder","volume-title":"2017 IEEE Winter Conference on Applications of Computer Vision (WACV)","author":"Hou","year":"2016"},{"key":"B18","doi-asserted-by":"crossref","DOI":"10.1609\/aaai.v32i1.11867","article-title":"Multi-level variational autoencoder: Learning disentangled representations from grouped observations","volume-title":"Proceedings of the AAAI conference on artificial intelligence","author":"Bouchacourt","year":"2018"},{"key":"B19","first-page":"1","article-title":"Introvae: Introspective variational autoencoders for photographic image synthesis","author":"Huang","year":"2018","journal-title":"Neural Inf Process Syst (NeurIPS)"},{"key":"B20","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/IJCNN48605.2020.9206694","article-title":"Deep learning based unsupervised and semi-supervised classification for keratoconus","volume-title":"2020 international joint conference on neural networks (IJCNN)","author":"Hallett","year":"2020"},{"key":"B21","doi-asserted-by":"publisher","DOI":"10.1167\/tvst.11.8.30","article-title":"Development of cumulative order-preserving image transformation based variational autoencoder for anterior segment optical coherence tomography images","volume":"11","author":"Shon","year":"2022","journal-title":"Trans Vision Sci Technol"},{"key":"B22","doi-asserted-by":"publisher","first-page":"20586","DOI":"10.1038\/s41598-023-46903-5","article-title":"Computer-aided diagnosis of keratoconus through VAE-augmented images using deep learning","volume":"13","author":"Agharezaei","year":"2023","journal-title":"Sci Rep"},{"key":"B23","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41598-019-54653-6","article-title":"Estimating rates of progression and predicting future visual fields in glaucoma using a deep variational autoencoder","author":"Berchuck","year":"2019","journal-title":"Sci Rep"},{"key":"B24","article-title":"Retina-VAE: Variationally decoding the spectrum of macular disease","volume":"1907","author":"Odaibo","year":"2019","journal-title":"ArXiv"},{"key":"B25","doi-asserted-by":"crossref","DOI":"10.1145\/3292500.3330757","article-title":"Glaucoma progression prediction using retinal thickness via latent space linear regression","volume-title":"Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & Data mining","author":"Zheng","year":"2019"},{"key":"B26","doi-asserted-by":"publisher","DOI":"10.1016\/j.ogla.2020.01.001","article-title":"Improving the structure-function relationship in glaucomatous visual fields by using a deep learning-based noise reduction approach","volume":"3","author":"Asaoka","year":"2020","journal-title":"Ophthalmol Glaucoma"},{"key":"B27","article-title":"Assessing glaucoma in retinal fundus photographs using deep feature consistent variational autoencoders","volume":"1","author":"Mandal","year":"2021","journal-title":"ArXiv"},{"key":"B28","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.xops.2023.100423","article-title":"Efficacy of smoothing algorithms to enhance detection of visual field progression in glaucoma","volume":"4","author":"Mohammadzadeh","year":"2024","journal-title":"Ophthalmol Sci"},{"key":"B29","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1167\/tvst.13.1.13","article-title":"Visualization of optic nerve structural patterns in papilledema using deep learning variational autoencoders","volume":"13","author":"Wang","year":"2024","journal-title":"Trans Vision Sci Technol"},{"key":"B30","doi-asserted-by":"crossref","DOI":"10.1145\/2939672.2939785","article-title":"Xgboost: A scalable tree boosting system","volume-title":"Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining","author":"Chen","year":"2016"},{"key":"B31","doi-asserted-by":"publisher","first-page":"790","DOI":"10.1016\/j.ophtha.2024.01.011","article-title":"Ophthalmic and systemic factors of acute nonarteritic anterior ischemic optic neuropathy in the quark207 treatment trial","volume":"131","author":"Kupersmith","year":"2024","journal-title":"Ophthalmology"},{"key":"B32","doi-asserted-by":"crossref","unstructured":"Waskom\n              M\n            \n          \n          Seaborn: Statistical data visualization\n          \n          2022","DOI":"10.21105\/joss.03021"},{"key":"B33","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2003.819861","article-title":"Image quality assessment: From error visibility to structural similarity","volume":"13","author":"Wang","year":"2004","journal-title":"IEEE Trans Image Process"},{"key":"B34","doi-asserted-by":"publisher","first-page":"98","DOI":"10.1109\/MSP.2008.930649","article-title":"Mean squared error: Love it or leave it? A new look at signal fidelity measures","volume":"26","author":"Wang","year":"2009","journal-title":"IEEE Signal Process Magazine"},{"key":"B35","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR46437.2021.00437","article-title":"Soft-introvae: Analyzing and improving the introspective variational autoencoder","volume-title":"2021 IEEE\/CVF conference on computer vision and pattern recognition (CVPR).","author":"Daniel","year":"2021"},{"key":"B36","article-title":"Semi-automated 2D bruch's membrane shape analysis in papilledema using spectral-domain optical coherence tomography","author":"Wang","year":"2015","journal-title":"SPIE Med Imaging"},{"key":"B37","doi-asserted-by":"publisher","DOI":"10.1167\/iovs.16-21089","article-title":"Peripapillary retinal pigment epithelium layer shape changes from acetazolamide treatment in the idiopathic intracranial hypertension treatment trial","volume":"58","author":"Wang","year":"2017","journal-title":"Invest Ophthalmol Visual Sci"}],"container-title":["Frontiers in Ophthalmology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fopht.2024.1497848\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,2,3]],"date-time":"2025-02-03T06:36:22Z","timestamp":1738564582000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fopht.2024.1497848\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,2,3]]},"references-count":37,"alternative-id":["10.3389\/fopht.2024.1497848"],"URL":"https:\/\/doi.org\/10.3389\/fopht.2024.1497848","relation":{},"ISSN":["2674-0826"],"issn-type":[{"value":"2674-0826","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,2,3]]},"article-number":"1497848"}}