DGRNet: Disagreement-Guided Refinement for Uncertainty-Aware Brain Tumor Segmentation

Abstract

Accurate brain tumor segmentation from MRI scans is critical for diagnosis and treatment planning. Despite the strong performance of recent deep learning approaches, two fundamental limitations remain: (1) the lack of reliable uncertainty quantification in single-model predictions, which is essential for clinical deployment because the level of uncertainty may impact treatment decision-making, and (2) the under-utilization of rich information in radiology reports that can guide segmentation in ambiguous regions. In this paper, we propose the Disagreement-Guided Refinement Network (DGRNet), a novel framework that addresses both limitations through multi-view disagreement-based uncertainty estimation and text-conditioned refinement. DGRNet generates diverse predictions via four lightweight view-specific adapters attached to a shared encoder-decoder, enabling efficient uncertainty quantification within a single forward pass. Afterward, we build disagreement maps to identify regions of high segmentation uncertainty, which are then selectively refined according to clinical reports. Moreover, we introduce a diversity-preserving training strategy that combines pairwise similarity penalties and gradient isolation to prevent view collapse. The experimental results on the TextBraTS dataset show that DGRNet favorably improves state-of-the-art segmentation accuracy by 2.4% and 11% in main metrics Dice and HD95, respectively, while providing meaningful uncertainty estimates.

Figures (3)

• Figure 1: Overview of the DGRNet architecture. The model consists of four main components: (1) backbone extracts visual features via the Swin Transformer encoder, and textual features via the frozen BioBERT lee_2020_biobert, fused at the bottleneck, (2) view-specific adapters generate diverse predictions $Z^{(1)}$, $Z^{(2)}$, and $Z^{(3)}$ using learnable FiLM modulation with shared decoder and view-specific output heads (Sec. \ref{['sec::view_specific_adapters']}), (3) disagreement module computes an uncertainty map $U$ from variance, pairwise disagreement, and entropy of mean predictions with learnable weights (Sec. \ref{['sec::disagreement_module']}), and (4) refinement module uses disagreement-aware spatial attention and text-guided FiLM conditioning to produce a gated residual refinement $\Delta Z$, yielding the final prediction for all three sub-regions $\tilde{Z} = \hat{Z} + g \cdot \Delta Z$. (Sec. \ref{['sec::refinement_module']})
• Figure 2: Qualitative comparison with the baseline, using a representative sample that contains all of the sub-regions (TC, WT, and ET).
• Figure 3: The strong correlation between segmentation errors and the uncertainty map.