Precise and automated tissue segmentation plays a vital role in diagnosing glaucoma using ocular optical coherence tomography (OCT) images. The intricate anatomical structure of the peripapillary region, combined with the existence of the optic disc, contributes to the difficulty of the task. In order to fix this challenge, we have created a cutting-edge two-stage architecture that uses a convolutional graph network (GCN) to identify both the ocular disc and nine ocular layers concurrently. This method incorporates a multifaceted spatial processing module into a U-shaped neural network, leveraging anatomical previous knowledge to strengthen spatial thinking skills. The module is placed between the encoder and decoder portions of the network. Our segmentation network achieved a Dice score of 0.850±0.0012 and a pixel accuracy of 0.895±0.0019, outperforming existing state-of-the-art techniques.