An Attention-Guided Deep Regression Model for Landmark Detection in Cephalograms

Cephalometric tracing method is usually used in orthodontic diagnosis and treatment planning. In this paper, we propose a deep learning based framework to automatically detect anatomical landmarks in cephalometric X-ray images. We train the deep encoder-decoder model for landmark detection, which combines global landmark configuration with local high-resolution feature responses. The proposed framework is based on a 2-stage u-net, regressing the multi-channel heatmaps for landmark detection. In this framework, we embed attention mechanism with global stage heatmaps, guiding the local stage inferring, to regress the local heatmap patches in a high resolution. Besides, an Expansive Exploration strategy is applied to improve robustness while inferring, expanding the searching scope without increasing model complexity. We have evaluated the proposed framework in the most widely-used public dataset of landmark detection in cephalometric X-ray images. With less computation and manually tuning, the proposed framework achieves state-of-the-art results.