{ "display_name": "Synthesis: MR", "short_description": "Description:
Supervised MRI (T1-weighted) synthesis model developed as part of the SynthRAD 2025 Challenge (Task 1).

How to cite:
V. Boussot et al., Why Registration Quality Matters: Enhancing sCT Synthesis with IMPACT-Based Registration, arXiv preprint arXiv:2510.21358, https://doi.org/10.48550/arXiv.2510.21358.", "description": "Description:
Supervised MRI (T1-weighted) synthesis model developed as part of the SynthRAD 2025 Challenge (Task 1).

Architecture:
Based on a 2.5D UNet++ with a ResNet34 encoder, the model was optimized using the IMPACT-Synth loss, a perceptual loss leveraging semantic priors from SAM 2.1-s. Training was conducted with the KonfAI deep learning framework.

Training data:
Paired CT–MRI (T1-weighted) volumes from the SynthRAD 2025 Challenge (Task 1), aligned using IMPACT-based registration. Corresponding B-spline deformation fields are available in the SynthRAD2025-IMPACT (aligned) dataset repository.

How to cite:
V. Boussot et al., Why Registration Quality Matters: Enhancing sCT Synthesis with IMPACT-Based Registration, arXiv preprint arXiv:2510.21358, https://doi.org/10.48550/arXiv.2510.21358.", "tta": 2, "mc_dropout": false, "models": ["CV_0.pt", "CV_1.pt", "CV_2.pt", "CV_3.pt", "CV_4.pt"], "inputs": { "MR": { "display_name": "MR", "volume_type": "VOLUME", "required": true } }, "outputs": { "sCT": { "display_name": "sCT", "volume_type": "VOLUME", "required": true } }, "inputs_evaluations": { "Image": { "Evaluation.yml": { "sCT": { "display_name": "sCT", "volume_type": "VOLUME", "required": true }, "CT": { "display_name": "CT", "volume_type": "VOLUME", "required": true } } } } }