Improving Brain-to-Image Reconstruction via Fine-Grained Text Bridging