GAPPTANet: Geometry-Adaptive Physics-Prior Thermal Ablation Network for Rapid 3D Temperature Prediction During Microwave Ablation in Nonperfused Liver Tissue
Annals of Biomedical Engineering · 07.11.2026
DOI: 10.1007/s10439-026-04451-9 ↗Kurz gefasst
Fachartikel in der Zeitschrift Annals of Biomedical Engineering, online erschienen am 07.11.2026. Verfasst von Feifei Ding und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biomedical Engineering, College of Chemistry and Life Science, Beijing University of Technology, Beijing, 100124, China. Die Arbeit nennt 56 Literaturangaben. Gefördert von: Beijing Natural Science Foundation, Beijing Postdoctoral Research Foundation, R&D Program of the Beijing Municipal Education Commission.
Abstract (Originalsprache)
PURPOSE: Accurate three-dimensional temperature prediction is important for planning microwave ablation because incomplete heating may leave residual tumor, whereas excessive heating may damage adjacent structures. This study develops a fast and physically interpretable surrogate for predicting the 180-s endpoint temperature field in nonperfused liver tissue during single-antenna microwave ablation under a fixed protocol. METHODS: GAPPTANet introduces an analytical specific absorption rate (SAR) field as an explicit heat-source prior, applies a Geometry-Aware Adapter to correct applicator-induced near-field approximation errors, and uses a Gated Dual-Branch Decoder to reconstruct global heat diffusion and local high-temperature boundaries. The model was trained and evaluated using 27,050 high-fidelity finite-difference frequency-domain electromagnetic-thermal simulations. Performance was assessed using full-field errors, 50 and 60 ∘ C isotherm morphology, computational efficiency, ablation studies, and point-wise measurements from a limited ex vivo porcine liver experiment. RESULTS: GAPPTANet achieved a full-field root-mean-square error of 0.6001 ∘ C, a mean relative error of 0.5904%, and a batch-amortized GPU forward pass time of 1.58 ms per sample. It outperformed convolutional baselines and three neural-operator architectures under the same test setting. The ex vivo experiment yielded a pooled mean absolute error of 2.83 ± 1.82 ∘ C . CONCLUSION: Explicit heat-source priors combined with geometry-aware correction enable accurate, interpretable, and millisecond-scale 3D temperature prediction, supporting further development of AI-assisted microwave ablation planning.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Biomedical Engineering, College of Chemistry and Life Science, Beijing University of Technology, Beijing, 100124, China.
- College of Biomedical Engineering, Capital Medical University, Beijing, 100069, China.
- Liver Disease and Tumor Interventional Treatment Center, Beijing Youan Hospital, Capital Medical University, Beijing, 100069, China.
- Department of Biomedical Engineering, College of Chemistry and Life Science, Beijing University of Technology, Beijing, 100124, China. zhouzh@bjut.edu.cn.
- Department of Biomedical Engineering, College of Chemistry and Life Science, Beijing University of Technology, Beijing, 100124, China. wushuicai@bjut.edu.cn.
Förderung
- Beijing Natural Science Foundation
- Beijing Postdoctoral Research Foundation
- R&D Program of the Beijing Municipal Education Commission
Zitieren
APA
Ding, F., Wu, W., Wang, Q., Zhang, H., Siddiqui, S., Zhou, Z., & Wu, S. (2026). GAPPTANet: Geometry-Adaptive Physics-Prior Thermal Ablation Network for Rapid 3D Temperature Prediction During Microwave Ablation in Nonperfused Liver Tissue. Annals of Biomedical Engineering. https://doi.org/10.1007/s10439-026-04451-9
Vancouver
Ding F, Wu W, Wang Q, Zhang H, Siddiqui S, Zhou Z, et al. GAPPTANet: Geometry-Adaptive Physics-Prior Thermal Ablation Network for Rapid 3D Temperature Prediction During Microwave Ablation in Nonperfused Liver Tissue. Annals of Biomedical Engineering. 2026. doi:10.1007/s10439-026-04451-9
BibTeX
@article{Ding2026,
author = {Ding, Feifei and Wu, Weiwei and Wang, Qi and Zhang, Honghai and Siddiqui, Saima and Zhou, Zhuhuang and Wu, Shuicai},
title = {GAPPTANet: Geometry-Adaptive Physics-Prior Thermal Ablation Network for Rapid 3D Temperature Prediction During Microwave Ablation in Nonperfused Liver Tissue},
journal = {Annals of Biomedical Engineering},
year = {2026},
doi = {10.1007/s10439-026-04451-9}
}
EUVIMED