Improving PET/CT-Based Whole-Body Lesion Segmentation Using Prediction Uncertainty-Augmented Models
Quick Answer
This study introduces an uncertainty-aware framework for PET/CT lesion segmentation using nnU-Net, enhancing robustness through Bayesian ensembling and voxel-wise uncertainty quantification.
Quick Take
This study introduces an uncertainty-aware framework for PET/CT lesion segmentation using nnU-Net, enhancing robustness through Bayesian ensembling and voxel-wise uncertainty quantification. Evaluated on AutoPET-III and Deep-PSMA datasets, the model shows improved performance and lesion recovery, despite a precision-recall trade-off.
Key Points
- Bayesian ensembling reduces training stochasticity and improves model robustness.
- Uncertainty maps identify regions of model disagreement, correlating with misclassifications.
- Uncertainty-augmented training enhances lesion detection but increases false positives.
- The study uses two public datasets: AutoPET-III (1,611 scans) and Deep-PSMA (200 scans).
- First systematic investigation of uncertainty quantification in multi-tracer PET/CT segmentation.
Paper Resources
Article Content
From source RSS / original summaryarXiv:2606. 10115v1 Announce Type: new Abstract: Accurate lesion segmentation from whole-body Positron Emission Tomography (PET)/Computed Tomography (CT) scans is essential for cancer staging and treatment planning. PET provides functional metabolic information with different radiotracers, while CT offers anatomical localization. Lesion delineation from PET/CT imaging is clinically challenging due to subtle imaging features, confounders, and inter-reader variability.
Existing deep learning approaches suffer from training-related stochasticity, inconsistent predictions, missed lesions in high tumor-burden cases, and lack uncertainty quantification, limiting their clinical reliability.
Using nnU-Net as a baseline, we propose an uncertainty-aware framework for whole-body PET/CT lesion segmentation that integrates (1) Bayesian ensembling to reduce training stochasticity, (2) voxel-wise uncertainty quantification with epistemic and aleatoric decomposition, and (3) epistemic uncertainty-augmented training to improve lesion detection.
Two public datasets, AutoPET-III (1,611 scans) and Deep-PSMA (200 scans), comprising FDG and PSMA studies across multiple cancer types, are used for training and evaluation. Bayesian ensembling improves robustness and performance over deterministic nnU-Net models on the unseen AutoPET-III test set. Uncertainty maps highlight regions of model disagreement and correlate with misclassifications, particularly false positives.
Uncertainty-augmented training improves lesion recovery at the cost of increased FPVol, reflecting a precision-recall trade-off. A case-adaptive routing strategy further improves Dice by selecting between the base and augmented models. To our knowledge, this is the first study to systematically investigate uncertainty quantification in multi-tracer, pan-cancer PET/CT segmentation and to combine Bayesian ensembling with uncertainty-aware modeling for this task.
Reader Mode unavailable (could not extract clean content).
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →LLM-Guided ANN Index Optimization for Human-Object Interaction Retrieval
A phase-aware LLM agent optimizes human-object interaction retrieval, outperforming Optuna TPE by 33.3% and VDTuner by 34.2% on the HICO-DET benchmark. This method enhances throughput by 15.3x over UniIR and demonstrates strong transferability across vector database management systems.