Loading
Life sciences · Preprint
arXiv · September 8, 2026
Raises a question worth testing. It does not answer one.
SynthRCT is a computational method using conditional variational autoencoders to generate synthetic deformed CT images for robustness testing in proton therapy planning. The work is a technical proof-of-concept validated on respiratory 4DCT data, demonstrating feasibility of memory-scalable synthesis but providing no evidence of clinical utility, accuracy relative to ground truth, or impact on treatment planning decisions.
Methods validation study on retrospective imaging data. Respiratory 4DCT data with multiple breathing-phase anatomies per subject. Specific inclusion criteria, number of subjects, and institution not stated.. Intervention: SynthRCT: conditional variational autoencoder framework that learns latent deformation space and generates synthetic deformed CT images via local stationary velocity fields..
SynthRCT generates patient-specific anatomical deformations by learning a latent deformation space from multi-phase 4DCT data Local stationary velocity fields are assembled into coherent full-volume transformations for large field-of-view CT imaging Approach enables sampling beyond predefined robustness scenarios, offering scalable memory usage for synthetic repeat CT generation
Safety was not reported in the material analysed. Check the source before drawing any conclusion about harm.
If validated clinically, this method could enable more realistic anatomical perturbation scenarios for proton therapy robustness evaluation. However, no clinical validation, comparison with current planning methods, or outcome data is provided.
This is a methodological validation study of a machine learning tool for synthetic medical image generation, not a clinical trial or outcomes study; it demonstrates technical feasibility on imaging data without clinical endpoint evaluation.
As stated by the source record.
If validated clinically, this method could enable more realistic anatomical perturbation scenarios for proton therapy robustness evaluation. However, no clinical validation, comparison with current planning methods, or outcome data is provided.
Graded across the dimensions that decide whether you should act, each from what the source actually supports. There is no single score, and where a dimension was not assessed it says so.
What is missing. This record has no reported figures. That is a gap in the analysis, not a judgement about the study.
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.
Taken from the source record, never inferred. Follow any of these and new work involving them reaches your briefing.