The field of ophthalmic image analysis is rapidly advancing with the development of innovative deep learning approaches. Recent research has focused on improving the accuracy and efficiency of image segmentation and disease detection models. Notably, attention mechanisms and optimized convolutional layers have been incorporated into existing architectures to enhance performance. Furthermore, efforts have been made to create universal representations and instant performance prediction for neural architectures, allowing for more flexible and scalable models. Benchmarking initiatives have also been established to systematically evaluate the performance of various models across different datasets and tasks. Overall, these advancements have the potential to improve the diagnosis and treatment of ophthalmic diseases. Noteworthy papers include: ONNX-Net, which presents a universal representation for neural architectures, and U-Bench, a comprehensive benchmark for evaluating U-Net variants. Additionally, the paper on detecting retinal diseases using an accelerated reused convolutional network demonstrates improved accuracy and efficiency in disease detection.