Neural Network-Driven Direct CBCT-Based Dose Calculation for Head-and-Neck Proton Treatment Planning

2025-09-22

AI Summary

In head-and-neck proton therapy, poor CBCT image quality necessitates complex HU-to-stopping-power-ratio (SPR) calibration for accurate dose calculation, impeding clinical adoption of adaptive radiotherapy. To address this, we propose a direct CBCT-based proton dose prediction method that bypasses image-domain correction entirely. Our approach introduces a novel xLSTM-based architecture that jointly encodes energy tokens and beam-angle sequences, trained end-to-end using high-fidelity Monte Carlo–generated dose ground truth. Validated on five patient cases, the method achieves a gamma pass rate (3%/2 mm) of 95.1 ± 2.7% and a mean absolute error of 2.6 ± 1.4% in high-dose regions, with per-case computation time under three minutes. To our knowledge, this is the first CBCT-driven proton dose prediction framework that is fast, accurate, and calibration-free—substantially enhancing the clinical feasibility and efficiency of adaptive proton therapy.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Large Vision Models
• Cognitive Modeling & Cognitive Systems: Adaptive Behavior

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Accurate dose calculation on cone beam computed tomography (CBCT) images is essential for modern proton treatment planning workflows, particularly when accounting for inter-fractional anatomical changes in adaptive treatment scenarios. Traditional CBCT-based dose calculation suffers from image quality limitations, requiring complex correction workflows. This study develops and validates a deep learning approach for direct proton dose calculation from CBCT images using extended Long Short-Term Memory (xLSTM) neural networks. A retrospective dataset of 40 head-and-neck cancer patients with paired planning CT and treatment CBCT images was used to train an xLSTM-based neural network (CBCT-NN). The architecture incorporates energy token encoding and beam's-eye-view sequence modelling to capture spatial dependencies in proton dose deposition patterns. Training utilized 82,500 paired beam configurations with Monte Carlo-generated ground truth doses. Validation was performed on 5 independent patients using gamma analysis, mean percentage dose error assessment, and dose-volume histogram comparison. The CBCT-NN achieved gamma pass rates of 95.1 $pm$ 2.7% using 2mm/2% criteria. Mean percentage dose errors were 2.6 $pm$ 1.4% in high-dose regions ($>$90% of max dose) and 5.9 $pm$ 1.9% globally. Dose-volume histogram analysis showed excellent preservation of target coverage metrics (Clinical Target Volume V95% difference: -0.6 $pm$ 1.1%) and organ-at-risk constraints (parotid mean dose difference: -0.5 $pm$ 1.5%). Computation time is under 3 minutes without sacrificing Monte Carlo-level accuracy. This study demonstrates the proof-of-principle of direct CBCT-based proton dose calculation using xLSTM neural networks. The approach eliminates traditional correction workflows while achieving comparable accuracy and computational efficiency suitable for adaptive protocols.

Problem

• Accurate proton dose calculation on CBCT images for adaptive head-and-neck treatment planning
• Overcoming CBCT image quality limitations that require complex correction workflows
• Developing direct dose calculation method using xLSTM neural networks for efficiency

Innovation

• Uses xLSTM neural networks for direct CBCT dose calculation
• Incorporates energy token encoding for proton dose patterns
• Implements beam's-eye-view sequence modeling for spatial dependencies