Jui-Kai Wang ^1,2* Brett A Johnson ² Zhi Chen ³ Honghai Zhang ³ David Szanto ⁴ Brian Woods ^4,5 Michael Wall ² Young H Kwon ² Edward F Linton ² Andrew Pouw ² Mark J Kupersmith ^4,6,7 Mona K Garvin ^1,8 Randy H Kardon ^1,2

• ¹ Center for the Prevention and Treatment of Visual Loss, Iowa City VA Health Care System, Iowa City, Iowa, United States
• ² Department of Ophthalmology and Visual Sciences, Carver College of Medicine, The University of Iowa, Iowa City, United States
• ³ The Iowa Institute for Biomedical Imaging, The University of Iowa, Iowa City, Iowa, United States
• ⁴ Department of Ophthalmology, Icahn School of Medicine at Mount Sinai, New York, New York, United States
• ⁵ Galway University Hospital, Galway, Ireland
• ⁶ Department of Neurosurgery, Icahn School of Medicine at Mount Sinai, New York, New York, United States
• ⁷ Department of Neurology, Icahn School of Medicine at Mount Sinai, New York, New York, United States
• ⁸ Department of Electrical and Computer Engineering, College of Engineering, The University of Iowa, Iowa City, Iowa, United States

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.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.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 ± 2.29 µm (two dLVs only) to 4.02 ± 1.61 µm (two dLVs and eight bLVs). However, the image-based structural similarity index (SSIM) remained similar (0.91 ± 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.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's ability to capture key spatial features and predict disease progression. Future work will focus on integrating additional image modalities to further refine the model's diagnostic capabilities.