Search papers, labs, and topics across Lattice.
This paper introduces SHTA, a novel training-time semantic representation branch aimed at enhancing semantic consistency in hard regions of semi-supervised medical image segmentation. By refining intermediate semantic representations through techniques like Semantic Assignment and Hard Token Refinement, SHTA improves segmentation accuracy without adding inference costs. Evaluations on the Synapse and AMOS datasets show that SHTA consistently enhances performance across various segmentation frameworks, particularly in weak-organ recovery and reducing semantic ambiguity.
SHTA achieves significant improvements in segmentation accuracy by ensuring semantic consistency in challenging regions, all without increasing inference costs.
Recent advances in semi-supervised medical image segmentation have achieved remarkable performance through prediction consistency, pseudo-label supervision, and hard-region supervision. However, these methods primarily improve supervision quality rather than explicitly enforcing semantic consistency in the learned representations of hard regions. Consequently, even under increasingly stronger prediction-level supervision, difficult regions exhibiting unstable semantic assignment often fail to establish semantically consistent representations during training, thereby limiting further segmentation improvement. To address this issue, we propose SHTA (Semantic Hard Token Correction and Center Alignment), a lightweight training-time semantic representation branch. Instead of introducing additional prediction supervision, SHTA refines intermediate semantic representations through Semantic Assignment, Hard Token Refinement, and Semantic Center Alignment, thereby improving semantic consistency in hard regions while preserving the original prediction pathway and introducing no additional inference cost. We integrate SHTA into representative semi-supervised segmentation frameworks, including GA-CPS, CPS, URPC, and MagicNet, and conduct evaluations on the Synapse and AMOS datasets. Experimental results demonstrate that SHTA delivers consistent paired improvements across frameworks, with especially clear gains in segmentation accuracy, weak-organ recovery, and semantic ambiguity reduction, while incurring only training-time overhead. The code is available at https://anonymous.4open.science/r/release_SHTA-42D5/.
