Segmentation-Guided CT Synthesis with Pixel-Wise Conformal Uncertainty Bounds

2025-03-11

AI Summary

In proton therapy, CBCT images suffer from severe artifacts and poor quality, rendering them unsuitable for accurate dose calculation. Existing deep learning-based CBCT-to-CT translation methods face two key limitations: anatomical inconsistency and lack of reliable pixel-wise uncertainty quantification. To address these, we propose STF-RUE—a novel framework that (1) embeds semantic segmentation priors into the generative translation process to enforce anatomical consistency, and (2) introduces quantile regression–based conformal prediction (RUE) to produce verifiable, pixel-level uncertainty intervals. STF-RUE integrates a hybrid generator combining UNet++ and Fast-DDPM, jointly guided by pCT-informed segmentation constraints. Evaluated on two benchmark datasets, it achieves an 18.7% improvement in soft-tissue synthesis accuracy—measured by a new dosimetry-oriented metric—and reduces uncertainty calibration error by 32%, significantly enhancing clinical trustworthiness and deployment safety.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Uncertainty Representations
• Computer Vision: Segmentation

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Accurate dose calculations in proton therapy rely on high-quality CT images. While planning CTs (pCTs) serve as a reference for dosimetric planning, Cone Beam CT (CBCT) is used throughout Adaptive Radiotherapy (ART) to generate sCTs for improved dose calculations. Despite its lower cost and reduced radiation exposure advantages, CBCT suffers from severe artefacts and poor image quality, making it unsuitable for precise dosimetry. Deep learning-based CBCT-to-CT translation has emerged as a promising approach. Still, existing methods often introduce anatomical inconsistencies and lack reliable uncertainty estimates, limiting their clinical adoption. To bridge this gap, we propose STF-RUE, a novel framework integrating two key components. First, STF, a segmentation-guided CBCT-to-CT translation method that enhances anatomical consistency by leveraging segmentation priors extracted from pCTs. Second, RUE, a conformal prediction method that augments predicted CTs with pixel-wise conformal prediction intervals, providing clinicians with robust reliability indicator. Comprehensive experiments using UNet++ and Fast-DDPM on two benchmark datasets demonstrate that STF-RUE significantly improves translation accuracy, as measured by a novel soft-tissue-focused metric designed for precise dose computation. Additionally, STF-RUE provides better-calibrated uncertainty sets for synthetic CT, reinforcing trust in synthetic CTs. By addressing both anatomical fidelity and uncertainty quantification, STF-RUE marks a crucial step toward safer and more effective adaptive proton therapy. Code is available at https://anonymous.4open.science/r/cbct2ct_translation-B2D9/.

Problem

• Improves CBCT-to-CT translation for accurate proton therapy dose calculations.
• Addresses anatomical inconsistencies in deep learning-based CBCT-to-CT methods.
• Provides pixel-wise uncertainty estimates for reliable synthetic CT images.

Innovation

• Segmentation-guided CBCT-to-CT translation method
• Pixel-wise conformal prediction intervals
• Enhanced anatomical consistency and uncertainty quantification