# Eliminating Registration Bias in Synthetic CT Generation: A Physics-Based Simulation Framework

arXiv: https://arxiv.org/abs/2602.02130 · HTML: https://mlsanity.com/paper/2602.02130/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 5 of its 31 references that are in the map.
- First posted on arXiv 2 Feb 2026.

## Builds on (most cited first, 5 of 5)

- [DeepDRR -- A Catalyst for Machine Learning in Fluoroscopy-guided Procedures](https://mlsanity.com/paper/1803.08606/) (2018)
- [SynthRAD2023 Grand Challenge dataset: generating synthetic CT for radiotherapy](https://mlsanity.com/paper/2303.16320/) (2023)
- [SynthRAD2025 Grand Challenge dataset: generating synthetic CTs for radiotherapy](https://mlsanity.com/paper/2502.17609/) (2025)
- [Why Registration Quality Matters: Enhancing sCT Synthesis with IMPACT-Based Registration](https://mlsanity.com/paper/2510.21358/) (2025)
- [SinoSynth: A Physics-based Domain Randomization Approach for Generalizable CBCT Image Enhancement](https://mlsanity.com/paper/2409.18355/) (2024)

## Built on by (most cited first, 1 of 1)

- [When Misalignment Becomes Supervision: Structured Label Noise in Supervised Synthetic CT Generation](https://mlsanity.com/paper/2609.29387/) (2026)

## Abstract (arXiv)

Supervised synthetic CT generation from CBCT requires registered training pairs, yet perfect registration between separately acquired scans remains unattainable. This registration bias propagates into trained models and corrupts standard evaluation metrics. This may suggest that superior benchmark performance indicates better reproduction of registration artifacts rather than anatomical fidelity. We propose physics-based CBCT simulation to provide geometrically aligned training pairs by construction, combined with evaluation using geometric alignment metrics against input CBCT rather than biased ground truth. On two independent pelvic datasets, models trained on synthetic data achieved superior geometric alignment (Normalized Mutual Information: 0.31 vs 0.22) despite lower conventional intensity scores. Intensity metrics showed inverted correlations with clinical assessment for deformably registered data, while Normalized Mutual Information consistently predicted observer preference across registration methodologies (rho = 0.31, p < 0.001). Clinical observers preferred synthetic-trained outputs in 87% of cases, demonstrating that geometric fidelity, not intensity agreement with biased ground truth, aligns with clinical requirements.