Stable training and confidence‑calibrated generation of urinary stone CT images via physics‑constrained GANs under small‑sample conditions
Biomedical Physics & Engineering Express · 09.11.2026
DOI: 10.1088/2057-1976/aebfa4 ↗Kurz gefasst
Fachartikel in der Zeitschrift Biomedical Physics & Engineering Express, online erschienen am 09.11.2026. Verfasst von Jiangfei Wang und 6 weiteren Autor:innen. Beteiligte Einrichtung: Shihezi University, shihezi University, Shihezi, Xinjiang, 832003, China. Gefördert von: Innovation Leading Talent Project of Prefecture-level Cities of the Xinjiang Production and Construction Corps, "Tianshan Talent" cultivation program for high-level talent training, This paper was supported by National Natural Science Foundation of China.
Abstract (Originalsprache)
In recent years, artificial intelligence has become integral to modern medicine. Deep learning algorithms are widely used to assist clinical diagnosis, and medical image synthesis has become increasingly mature. However, in small-sample scenarios, synthetic images often overlook infrequent but clinically valuable lesion features. To address the challenges of training instability, insufficient authenticity, and limited clinical usability, we propose a two-stage framework based on Generative Adversarial Networks (GANs). In the first stage, we establish a stable training and physical prior constraint framework for small-sample CT images. Specifically, we incorporate an established asymmetric normalization strategy into the WGAN-GP framework to balance generator convergence and the discriminator's Lipschitz constraint, and we embed a physical noise simulation layer at the discriminator's front end, combined with partial volume effect gradient regularization, to enhance the realism of synthesized images. In the second stage, we employ MC Dropout and deep ensemble methods to quantify uncertainty, generating uncertainty heatmaps and physically consistent confidence maps based on the Poisson-Gaussian noise model. This assists physicians in locating abnormal areas and reducing misdiagnosis risks. Through joint optimization of the generation framework, uncertainty quantification, and confidence image output, the proposed framework effectively resolves issues of mode collapse, low confidence, and poor clinical usability under small-sample conditions. The proposed framework achieves a competitive FID of 128.0, an Expected Calibration Error (ECE) of 0.072, and a high-risk recall rate of 88% for stone boundary regions, demonstrating its potential for clinically reliable synthetic CT generation.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Shihezi University, shihezi University, Shihezi, Xinjiang, 832003, China.
- Shihezi University, Shihezi University, Shihezi, Xinjiang, 832003, China.
- The General Hospital of the Third Division of Xinjiang Production and Construction Corps, No. 13, Qianhai East Street, Tumxuk City, Tumxuk, 844000, China.
- Shihezi University, shihezi University, Shihezi, 832003, China.
ORCID-Profile: Liu Na
Förderung
- Innovation Leading Talent Project of Prefecture-level Cities of the Xinjiang Production and Construction Corps
- "Tianshan Talent" cultivation program for high-level talent training
- This paper was supported by National Natural Science Foundation of China
Zitieren
APA
Wang, J., Xie, X., Wang, Y., Liu, Y., Ma, Y., Lian, W., & Na, L. (2026). Stable training and confidence‑calibrated generation of urinary stone CT images via physics‑constrained GANs under small‑sample conditions. Biomedical Physics & Engineering Express. https://doi.org/10.1088/2057-1976/aebfa4
Vancouver
Wang J, Xie X, Wang Y, Liu Y, Ma Y, Lian W, et al. Stable training and confidence‑calibrated generation of urinary stone CT images via physics‑constrained GANs under small‑sample conditions. Biomedical Physics & Engineering Express. 2026. doi:10.1088/2057-1976/aebfa4
BibTeX
@article{Wang2026,
author = {Wang, Jiangfei and Xie, Xinrong and Wang, Yaohui and Liu, Yuanhao and Ma, Yongjie and Lian, Wenyong and Na, Liu},
title = {Stable training and confidence‑calibrated generation of urinary stone CT images via physics‑constrained GANs under small‑sample conditions},
journal = {Biomedical Physics & Engineering Express},
year = {2026},
doi = {10.1088/2057-1976/aebfa4}
}
EUVIMED