RecallRefine: Coverage-Constrained Coarse-to-Fine Segmentationfor Small 3D Lesions
- 1Zhengzhou University
Abstract
Coarse to fine segmentation is a natural strategy for three dimensional lesion delineation becausefull resolution processing of an entire scan is expensive while only a small fraction of the volume contains pathology. Its main failure mode, however, is structural: a lesion missed by the coarse stage is usually absent from the region of interest and therefore cannot be recovered by the fine stage. This problem is most severe for small, low contrast, and multifocal lesions. We introduce RECALLREFINE, a coverage constrained coarse to fine framework that treats candidate-region selection as a recall critical learning problem rather than a deterministic crop around the coarse mask.A full-volume network first predicts coarse lesion probabilities and multi-scale features. A candidate score network then combines coarse probability, predictive uncertainty, feature novelty, and spatial context to score overlapping 3D blocks. During training, a differentiable selector is optimized with a component-wise coverage objective that requires every annotated lesion instance to remain covered by at least one selected block under a fixed refinement budget. Selected blocks are processed at native resolution by a local refiner with a context-ring branch, and overlapping predictions are merged back into the full volume. Across four public 3D lesion benchmarks, RECALLREFINE improves mean tumor Dice from 76.5% for the strongest matched coarse-to-fine baseline to 78.4%, while increasing sensitivity for lesions below 10 mm from 60.9% to 70.8%.At a 20% candidate budget, the selector covers 94.4% of lesion instances, substantially reducing irreversible coarse-stage misses. The results suggest that the central problem in cascaded lesion segmentation is not only how to refine a detected lesion, but how to guarantee that difficult lesions are still represented in the refinement set.
Related
- Recent advances in coarse-grained simulations of nucleic acid condensation
- From coarse-grained molecular dynamics simulations to continuum models of biomolecular condensation
- Developing Explicit Base Stacking Potentials for the IsRNA2+ Coarse-Grained RNA Force Field Using Iterative Reweighting.
- Light Martini water accelerates sampling in coarse-grained molecular dynamics simulations
- Cross-taxon Coarse-Grained IDP Simulations Enable Architecture-Independent Generalization
BibTeX
@misc{chen2026recallrefine,
title = {{RecallRefine: Coverage-Constrained Coarse-to-Fine Segmentationfor Small 3D Lesions}},
author = {W. Chen and J. Chen},
howpublished = {bioRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.21.753396},
url = {https://www.biorxiv.org/content/10.64898/2026.09.21.753396v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - RecallRefine: Coverage-Constrained Coarse-to-Fine Segmentationfor Small 3D Lesions
AU - W. Chen
AU - J. Chen
PY - 2026
DA - 2026/09/21
PB - bioRxiv
DO - 10.64898/2026.09.21.753396
UR - https://www.biorxiv.org/content/10.64898/2026.09.21.753396v1
AB - Coarse to fine segmentation is a natural strategy for three dimensional lesion delineation becausefull resolution processing of an entire scan is expensive while only a small fraction of the volume contains pathology. Its main failure mode, however, is structural: a lesion missed by the coarse stage is usually absent from the region of interest and therefore cannot be recovered by the fine stage. This problem is most severe for small, low contrast, and multifocal lesions. We introduce RECALLREFINE, a coverage constrained coarse to fine framework that treats candidate-region selection as a recall critical learning problem rather than a deterministic crop around the coarse mask.A full-volume network first predicts coarse lesion probabilities and multi-scale features. A candidate score network then combines coarse probability, predictive uncertainty, feature novelty, and spatial context to score overlapping 3D blocks. During training, a differentiable selector is optimized with a component-wise coverage objective that requires every annotated lesion instance to remain covered by at least one selected block under a fixed refinement budget. Selected blocks are processed at native resolution by a local refiner with a context-ring branch, and overlapping predictions are merged back into the full volume. Across four public 3D lesion benchmarks, RECALLREFINE improves mean tumor Dice from 76.5% for the strongest matched coarse-to-fine baseline to 78.4%, while increasing sensitivity for lesions below 10 mm from 60.9% to 70.8%.At a 20% candidate budget, the selector covers 94.4% of lesion instances, substantially reducing irreversible coarse-stage misses. The results suggest that the central problem in cascaded lesion segmentation is not only how to refine a detected lesion, but how to guarantee that difficult lesions are still represented in the refinement set.
ER -