Conditional generative model creates realistic anatomical changes for proton therapy
SynthRCT: Scalable Conditional Deformation Synthesis for Synthetic Repeat CT Generation
Computer Vision and Pattern RecognitionArtificial IntelligenceMachine Learning
Summary
Proton therapy for cancer treatment needs careful planning based on medical images of a patient’s body, but bodies can change in complex ways during treatment. The authors developed SynthRCT, a method that can generate realistic 3D models showing how a patient’s anatomy might change, using advanced machine learning. This allows doctors to better test how treatment plans hold up under different possible body changes. Their approach works efficiently on large 3D scans and captures patient-specific variations beyond simple assumptions.
proton therapyCT scananatomical deformationconditional variational autoencoderlatent spacelocal stationary velocity field4DCTrobustness evaluationgenerative model
Authors
Tomas Guija-Valiente, Blanca Rodriguez-Gonzalez, Norberto Malpica
Abstract
In proton therapy, plans are typically optimized on a single planning CT, making robustness evaluation essential under anatomical changes. However, current scenarios often rely on simplified perturbations that poorly capture complex, patient-specific variability. We propose SynthRCT, a scalable conditional generative framework for 3D anatomical deformation synthesis. Based on a conditional variational autoencoder, SynthRCT learns a latent deformation space and decodes sampled latent codes into local stationary velocity fields conditioned on an input anatomy. Local fields are assembled into coherent full-volume transformations, enabling memory-scalable generation for large field-of-view CT data. We validate the approach on respiratory 4DCT data with multiple breathing-phase anatomies per subject. SynthRCT enables patient-specific sampling of plausible anatomical transformations beyond predefined robustness scenarios. Code available at: https://github.com/TomasGuija/SynthRCT.