To enhance the precision and reliability of Optic Cup and Optic Disc Segmentation, we present a novel Multi-scale Channel Attention Vision LSTM (MCA-ViLSTM) Network. It integrates the encoder-decoder framework of U-Net with the sequential modeling capabilities of Long Short-Term Memory (LSTM) networks, thereby enabling the model to simultaneously capture local morphological features and long-range spatial dependencies while maintaining computational efficiency. Furthermore, we introduce a Multi-scale Channel Attention (MCA) module that enhances feature representation for small anatomical structures through adaptive channel weighting mechanisms, effectively addressing segmentation challenges arising from class imbalance. Comprehensive experiments conducted on publicly available datasets (DRIONS-DB and REFUGE) demonstrate that MCA-ViLSTM achieves superior segmentation performance in fundus image analysis.