Spectral Domain Optical Coherence Tomography (SD-OCT) is a demanding imaging technique by which diagnosticians detect retinal diseases. Automating the procedure for early detection and diagnosis of retinal diseases has been proposed in many intricate ways through the use of image processing, machine learning and deep learning algorithms. Unfortunately, the traditional methods are erroneous in nature and quite expensive as they require additional participation from the human diagnosticians. In this chapter, we propose a comprehensive sets novel blocks for building a deep learning architecture to effectively differentiate between different pathologies causing retinal degeneration. We further show how integrating these novel blocks within a novel network architecture gives a better classification accuracy of these disease and addresses the preexisting problems with gradient explosion in the deep residual architectures. The technique proposed in this chapter achieves better accuracy compared to the state-of-the-art for two separately hosted Retinal OCT image data-set. Furthermore, we illustrate a real-time prediction system that by exploiting this deep residual architecture, consisting one of these novel blocks, outperforms expert ophthalmologists.