PEEL: Physics-Enabled Evidential Learning for Identifiable Uncertainty in CT Imaging
Abstract
Normal-inverse-gamma (NIG) regression is not uniquely identifiable from its marginal Student-t likelihood: the likelihood determines three combinations of four NIG parameters and is constant along a one-dimensional fiber. We identify that fiber using independent physical measurement. As an initial embodiment, a reconstruction network receives one noisy filtered-backprojection (FBP) image and is first trained only by Student-t negative log-likelihood to estimate the three identifiable coordinates (gamma, alpha, c). The network is then frozen; repeated physical-noise realizations propagated through its reconstruction output form a Monte Carlo (MC) teacher label for output-domain aleatoric variance. An aleatoric head attached to frozen features learns this label, after which (beta, nu) are recovered algebraically. On 30 held-out simulated objects at five photon levels, one-image predictions achieved pooled Spearman correlations of 0.832-0.951 against independent 400-repeat references, median within-image correlations were 0.834-0.947, and 98.81-99.55% of evaluated pixels satisfied the algebraic admissibility condition. The method needs no KL term, reference prior, evidence regularizer, or cross-loss weight.
Related
- Tangent-Plane Evidential Uncertainty in Active Learning for Magnetic Interatomic Potentials
- Physics-informed Learning for Orbital Uncertainty Propagation with Error Bounds
- Physics-informed random neural architectures on compact boundaryless multiply connected manifolds for random-operator models in machine learning and uncertainty quantification
- A physics-informed uncertainty-aware surrogate for p-n junction diodes
- Derivation of Novel Isotropic Metamaterials at the Hashin Shtrikman Stiffness Limit Enabled by Physics Informed Neural Networks
BibTeX
@misc{wang2026peel,
title = {{PEEL: Physics-Enabled Evidential Learning for Identifiable Uncertainty in CT Imaging}},
author = {Ge Wang},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.29599},
eprint = {2609.29599},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.29599},
}
RIS
TY - UNPB
TI - PEEL: Physics-Enabled Evidential Learning for Identifiable Uncertainty in CT Imaging
AU - Ge Wang
PY - 2026
DA - 2026/09/01
PB - arXiv
DO - 10.48550/arxiv.2609.29599
UR - https://arxiv.org/abs/2609.29599
AB - Normal-inverse-gamma (NIG) regression is not uniquely identifiable from its marginal Student-t likelihood: the likelihood determines three combinations of four NIG parameters and is constant along a one-dimensional fiber. We identify that fiber using independent physical measurement. As an initial embodiment, a reconstruction network receives one noisy filtered-backprojection (FBP) image and is first trained only by Student-t negative log-likelihood to estimate the three identifiable coordinates (gamma, alpha, c). The network is then frozen; repeated physical-noise realizations propagated through its reconstruction output form a Monte Carlo (MC) teacher label for output-domain aleatoric variance. An aleatoric head attached to frozen features learns this label, after which (beta, nu) are recovered algebraically. On 30 held-out simulated objects at five photon levels, one-image predictions achieved pooled Spearman correlations of 0.832-0.951 against independent 400-repeat references, median within-image correlations were 0.834-0.947, and 98.81-99.55% of evaluated pixels satisfied the algebraic admissibility condition. The method needs no KL term, reference prior, evidence regularizer, or cross-loss weight.
ER -