Why Registration Quality Matters: Enhancing sCT Synthesis with IMPACT-Based Registration

• URL: http://arxiv.org/abs/2510.21358v1
• Date: Fri, 24 Oct 2025 11:40:21 GMT
• Title: Why Registration Quality Matters: Enhancing sCT Synthesis with IMPACT-Based Registration
• Authors: Valentin Boussot, Cédric Hémon, Jean-Claude Nunes, Jean-Louis Dillenseger,
• Abstract summary: Our model is a 2.5D U-Net++ with a ResNet-34 encoder, trained jointly across anatomical regions and fine-tuned per region.<n>On the local test sets, IMPACT-based registration achieved more accurate and anatomically consistent alignments than mutual-information-based registration.
• Score: 1.2560645967579729
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We participated in the SynthRAD2025 challenge (Tasks 1 and 2) with a unified pipeline for synthetic CT (sCT) generation from MRI and CBCT, implemented using the KonfAI framework. Our model is a 2.5D U-Net++ with a ResNet-34 encoder, trained jointly across anatomical regions and fine-tuned per region. The loss function combined pixel-wise L1 loss with IMPACT-Synth, a perceptual loss derived from SAM and TotalSegmentator to enhance structural fidelity. Training was performed using AdamW (initial learning rate = 0.001, halved every 25k steps) on patch-based, normalized, body-masked inputs (320x320 for MRI, 256x256 for CBCT), with random flipping as the only augmentation. No post-processing was applied. Final predictions leveraged test-time augmentation and five-fold ensembling. The best model was selected based on validation MAE. Two registration strategies were evaluated: (i) Elastix with mutual information, consistent with the challenge pipeline, and (ii) IMPACT, a feature-based similarity metric leveraging pretrained segmentation networks. On the local test sets, IMPACT-based registration achieved more accurate and anatomically consistent alignments than mutual-information-based registration, resulting in improved sCT synthesis with lower MAE and more realistic anatomical structures. On the public validation set, however, models trained with Elastix-aligned data achieved higher scores, reflecting a registration bias favoring alignment strategies consistent with the evaluation pipeline. This highlights how registration errors can propagate into supervised learning, influencing both training and evaluation, and potentially inflating performance metrics at the expense of anatomical fidelity. By promoting anatomically consistent alignment, IMPACT helps mitigate this bias and supports the development of more robust and generalizable sCT synthesis models.

Related papers

• Do Generative Metrics Predict YOLO Performance? An Evaluation Across Models, Augmentation Ratios, and Dataset Complexity [43.338311770275745]
  We present a controlled evaluation of synthetic augmentation for YOLOv11 across three single-class detection regimes.<n>We benchmark six GAN-, diffusion-, and hybrid-based generators over augmentation ratios from 10% to 150% of the real training split.<n>For each dataset-generator-augmentation configuration, we compute pre-training dataset metrics under a matched-size bootstrap protocol.
  arXiv  Detail & Related papers  (2026-02-20T03:02:36Z)
• Eliminating Registration Bias in Synthetic CT Generation: A Physics-Based Simulation Framework [0.8699280339422538]
  Supervised synthetic generation from CBCT requires registered training pairs, yet perfect registration between acquired scans remains unattainable.<n>This registration bias propagates into trained models and corrupts standard evaluation metrics.<n>We propose physics-based CBCT simulation to provide geometrically aligned training pairs by construction.
  arXiv  Detail & Related papers  (2026-02-02T14:14:47Z)
• Fine-tuning Segment Anything for Real-Time Tumor Tracking in Cine-MRI [1.2560645967579729]
  We address the challenge of real-time tumor tracking in cine-MRI sequences of the thoracic and abdominal regions under strong data scarcity constraints.<n>Two complementary strategies were explored: (i) unsupervised registration with the IMPACT similarity metric and (ii) foundation model-based segmentation leveraging SAM 2.1.<n>The final model was selected based on the highest Dice Similarity Coefficient achieved on the validation set after fine-tuning.
  arXiv  Detail & Related papers  (2025-10-29T21:57:12Z)
• TCIP: Threshold-Controlled Iterative Pyramid Network for Deformable Medical Image Registration [21.283219565079413]
  We propose the Feature-Enhanced Residual Module (FERM) as the core component of each decoding layer in the pyramid network.<n>FERM comprises three sequential blocks that extract anatomical semantic features, learn to suppress irrelevant features, and estimate the final deformation field.<n>We coin the model that integrates FERM and TCI as Threshold-Controlled Iterative Pyramid (TCIP)
  arXiv  Detail & Related papers  (2025-10-09T01:38:40Z)
• Deep Learning-Based Cross-Anatomy CT Synthesis Using Adapted nnResU-Net with Anatomical Feature Prioritized Loss [3.583510367091671]
  We present a patch-based 3D nnUNet adaptation for MR to CT and CBCT to CT image translation using the multicenter SynthRAD2025 dataset.
  arXiv  Detail & Related papers  (2025-09-26T14:22:15Z)
• EqDiff-CT: Equivariant Conditional Diffusion model for CT Image Synthesis from CBCT [43.92108185590778]
  Cone-beam computed tomography (CBCT) is widely used for imageguided radiotherapy (IGRT)<n>We propose a novel diffusion-based conditional generative model, coined EqDiff-CT, to synthesize high-quality CT images from CBCT.
  arXiv  Detail & Related papers  (2025-09-26T05:51:59Z)
• Consistency Trajectory Models: Learning Probability Flow ODE Trajectory of Diffusion [56.38386580040991]
  Consistency Trajectory Model (CTM) is a generalization of Consistency Models (CM) CTM enables the efficient combination of adversarial training and denoising score matching loss to enhance performance. Unlike CM, CTM's access to the score function can streamline the adoption of established controllable/conditional generation methods.
  arXiv  Detail & Related papers  (2023-10-01T05:07:17Z)
• GSMorph: Gradient Surgery for cine-MRI Cardiac Deformable Registration [62.41725951450803]
  Learning-based deformable registration relies on weighted objective functions trading off registration accuracy and smoothness of the field. We construct a registration model based on the gradient surgery mechanism, named GSMorph, to achieve a hyper parameter-free balance on multiple losses. Our method is model-agnostic and can be merged into any deep registration network without introducing extra parameters or slowing down inference.
  arXiv  Detail & Related papers  (2023-06-26T13:32:09Z)
• Spatiotemporal Feature Learning Based on Two-Step LSTM and Transformer for CT Scans [2.3682456328966115]
  We propose a novel, effective, two-step-wise approach to tickle this issue for COVID-19 symptom classification thoroughly. First, the semantic feature embedding of each slice for a CT scan is extracted by conventional backbone networks. Then, we proposed a long short-term memory (LSTM) and Transformer-based sub-network to deal with temporal feature learning.
  arXiv  Detail & Related papers  (2022-07-04T16:59:05Z)
• Weakly-supervised Biomechanically-constrained CT/MRI Registration of the Spine [72.85011943179894]
  We propose a weakly-supervised deep learning framework that preserves the rigidity and the volume of each vertebra while maximizing the accuracy of the registration. We specifically design these losses to depend only on the CT label maps since automatic vertebra segmentation in CT gives more accurate results contrary to MRI. Our results show that adding the anatomy-aware losses increases the plausibility of the inferred transformation while keeping the accuracy untouched.
  arXiv  Detail & Related papers  (2022-05-16T10:59:55Z)
• Voxelmorph++ Going beyond the cranial vault with keypoint supervision and multi-channel instance optimisation [8.88841928746097]
  Recent Learn2Reg benchmark shows single-scale U-Net architectures fall short of state-of-the-art performance for abdominal or intra-patient lung registration. Here, we propose two straightforward steps that greatly reduce this gap in accuracy. First, we employ keypoint self-supervision with a novel network head that predicts a discretised heatmap. Second, we replace multiple learned fine-tuning steps by a single instance with hand-crafted features and the Adam optimiser.
  arXiv  Detail & Related papers  (2022-02-28T19:23:29Z)
• InDuDoNet+: A Model-Driven Interpretable Dual Domain Network for Metal Artifact Reduction in CT Images [53.4351366246531]
  We construct a novel interpretable dual domain network, termed InDuDoNet+, into which CT imaging process is finely embedded. We analyze the CT values among different tissues, and merge the prior observations into a prior network for our InDuDoNet+, which significantly improve its generalization performance.
  arXiv  Detail & Related papers  (2021-12-23T15:52:37Z)
• Learning Multi-Modal Volumetric Prostate Registration with Weak Inter-Subject Spatial Correspondence [2.6894568533991543]
  We introduce an auxiliary input to the neural network for the prior information about the prostate location in the MR sequence. With weakly labelled MR-TRUS prostate data, we showed registration quality comparable to the state-of-the-art deep learning-based method.
  arXiv  Detail & Related papers  (2021-02-09T16:48:59Z)
• CovidDeep: SARS-CoV-2/COVID-19 Test Based on Wearable Medical Sensors and Efficient Neural Networks [51.589769497681175]
  The novel coronavirus (SARS-CoV-2) has led to a pandemic. The current testing regime based on Reverse Transcription-Polymerase Chain Reaction for SARS-CoV-2 has been unable to keep up with testing demands. We propose a framework called CovidDeep that combines efficient DNNs with commercially available WMSs for pervasive testing of the virus.
  arXiv  Detail & Related papers  (2020-07-20T21:47:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.