aiqbench analysis
This research introduces Physics-Enabled Evidential Learning (PEEL) to resolve parameter identifiability issues in CT imaging by integrating physical measurements with deep learning models. The approach successfully estimates aleatoric uncertainty using a frozen network and Monte Carlo propagation, achieving high correlation with reference data without relying on standard regularization terms like KL divergence. By leveraging physics-informed constraints, the method provides more robust uncertainty quantification for medical imaging tasks where statistical likelihoods alone are insufficient.
PEEL: Physics-Enabled Evidential Learning for Identifiable Uncertainty in CT Imaging
arXiv:2609.29599v1 Announce Type: new 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.
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.
10 pages, 3 figures, 1 table. Proof-of-concept study of physics-enabled identification of NIG uncertainty in CT imaging
Artificial Intelligence (cs.AI); Medical Physics (physics.med-ph)
arXiv:2609.29599 [cs.AI]
(or arXiv:2609.29599v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2609.29599 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 271 words
Related coverage
The Machines that Make the Machines
NVIDIA · Oct 7, 2026
The Entangled Dimensions of Decision Theory
LessWrong · Jul 30, 2026
ELECTRIC: Evidential Learning-Enhanced CT Reconstruction via Iterative Correction
arXiv · Aug 5, 2026
Award-Winning Researcher Trains Robots to Make Educated Guesses
IEEE Spectrum · Jun 12, 2026