A Systematic Literature Review: Leveraging Vision Transformers for Glaucoma Diagnosis

Glaucoma is a condition that causes gradual damage to the optic nerve due to increased pressure inside the eye, impacting the patient's vision. The detection and monitoring of glaucoma rely on clinical observations before making a diagnosis or decision about the patient. This process involves comprehensive eye examinations, additional testing, and extensive data collection. However, manual analysis by ophthalmologists can be tedious, time-consuming, and prone to diagnostic errors. Therefore, developing techniques to process and analyze existing clinical data to detect glaucoma as early as possible is crucial. This study explores using vision transformers (ViT) in glaucoma detection and leveraging available data for this purpose. The study aims to provide a comprehensive understanding of ViT usage in segmentation, classification, and glaucoma detection, as well as to evaluate the effectiveness and limitations of this approach. The study results indicate that ViT performs well in analyzing images to detect glaucoma, with several development techniques yielding advanced results. However, this research has limitations, such as limited datasets and inconsistent evaluation metrics. Hence, we require more extensive and diverse datasets and thorough external validation to confirm the dependability and clinical usefulness of glaucoma detection methods using ViT.