Abstract
Deep learning has achieved remarkable performance for tumor segmentation, yet high overlap scores do not necessarily imply reliable boundary delineation or complete identification of disconnected tumor foci. While uncertainty quantification methods provide voxel-wise confidence estimates, conformal prediction (CP) offers finite-sample coverage guarantees through calibrated prediction sets. Existing CP methods for image segmentation commonly rely on morphological dilation. We show that this construction becomes \emph{vacuous} in the presence of disconnected segmentation errors: missed tumor components saturate the calibrated score, producing prediction sets that preserve formal coverage while becoming operationally uninformative. We characterize this failure mode through the distinction between formal and operational coverage, and propose two complementary extensions. Instance-level CP separates boundary and structural errors, whereas Geodesic CP replaces isotropic dilation with confidence-guided expansion, and we combine both in a unified formulation. On BraTS 2021, across 100 calibration/test splits, the unified formulation eliminates vacuity while preserving split-conformal validity. The code is available at: https://github.com/robustml-eurecom/conformal-tumor-segmentation.git
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/UNSURE2026_020.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: Not Submitted
Link to Open Review
Open Review Page: https://openreview.net/profile?id=~Maria_A_Zuluaga1
BibTex
@InProceedings{VarLui_Beyond_MICCAISAT2026,
author = { Vargas, Luisa AND Rossi, Simone AND Zuluaga, Maria A.},
title = { { Beyond Morphological Dilation: Revisiting Conformal Prediction for 3D Tumor Segmentation } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17260},
month = {pending},
page = {pending}
}