Sampling-free Monte Carlo dropout approximation for single-pass uncertainty quantification in deep learning-based PET reconstructions
Physics in Medicine & Biology · 09.11.2026
DOI: 10.1088/1361-6560/aebfef ↗Kurz gefasst
Fachartikel in der Zeitschrift Physics in Medicine & Biology, online erschienen am 09.11.2026. Verfasst von Stephen Z Liu und 6 weiteren Autor:innen. Beteiligte Einrichtung: Siemens Medical Solutions USA, Inc., Knoxville, TN 37932, United States of America. Hauptthemen laut MEDLINE: Deep Learning, Monte Carlo Method, Image Processing, Computer-Assisted, Positron-Emission Tomography. Gefördert von: Siemens Medical Solutions USA.
Abstract (Originalsprache)
Objective: Medical image reconstructions using deep learning (DL) can rapidly generate high-quality images from degraded measurements, but their clinical adoption remains challenging, largely due to the lack of fast and reliable methods to quantify aleatoric (data-driven) and epistemic (model-driven) uncertainties on a per-prediction basis. We aim to develop a computationally efficient approach for estimating both uncertainties at the voxel level in DL-based image reconstruction within a single inference run.
Approach: We present a moment propagation-based framework that analytically approximates the conventional, multi-realization Monte Carlo (MC) dropout by propagating feature-map means and marginal variances through network layers, followed by a pre-calibrated correction table to compensate for covariance effects omitted in the approximation. This enables joint estimation of reconstruction, aleatoric uncertainty and epistemic uncertainty in a single forward pass. The proposed framework was implemented on top of FastPET -- a network model synthesizing PET reconstructions from time-of-flight histo-images -- and evaluated in a patient pilot study of 26 total-body PET/CT acquisitions (16 [18F]-FDG and 10 [18F]-PSMA cases).
Main Results: Proposed method achieved relatively strong agreement with the 50-pass MC approach, with absolute differences
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Siemens Medical Solutions USA, Inc., Knoxville, TN 37932, United States of America.
ORCID-Profile: Stephen Z Liu, Deepak Bharkhada, Yifan Zheng, Vladimir Y Panin, Paul Schleyer
Förderung
- Siemens Medical Solutions USA
Zitieren
APA
Liu, S. Z., Zhang, M., Bharkhada, D., Zheng, Y., Panin, V. Y., Schleyer, P., & Conti, M. (2026). Sampling-free Monte Carlo dropout approximation for single-pass uncertainty quantification in deep learning-based PET reconstructions. Physics in Medicine & Biology. https://doi.org/10.1088/1361-6560/aebfef
Vancouver
Liu SZ, Zhang M, Bharkhada D, Zheng Y, Panin VY, Schleyer P, et al. Sampling-free Monte Carlo dropout approximation for single-pass uncertainty quantification in deep learning-based PET reconstructions. Physics in Medicine & Biology. 2026. doi:10.1088/1361-6560/aebfef
BibTeX
@article{Liu2026,
author = {Liu, Stephen Z and Zhang, Mengxi and Bharkhada, Deepak and Zheng, Yifan and Panin, Vladimir Y and Schleyer, Paul and Conti, Maurizio},
title = {Sampling-free Monte Carlo dropout approximation for single-pass uncertainty quantification in deep learning-based PET reconstructions},
journal = {Physics in Medicine & Biology},
year = {2026},
doi = {10.1088/1361-6560/aebfef}
}
EUVIMED