Convolutional bi-directional learning and spatial enhanced attentions for lung tumor segmentation

•A novel channel enhanced module to extract contextual relationships between different deep image feature tensor channels by convolutional bi-directional GRU.•Region-level attention to distinguish the contribution of different local regions and associated features to the global learning process.•Integrating spatial and position dependencies by a new position enhanced self-attention mechanism.•The generalization ability of our new model is validated using multiple different 3D image segmentation backbones.