| | --- |
| | license: apache-2.0 |
| | tags: |
| | - medical-imaging |
| | - image-registration |
| | - torchscript |
| | - impact |
| | - pretrained |
| | - segmentation |
| | --- |
| | |
| | # 🧠 TorchScript Models for the IMPACT Semantic Similarity Metric |
| |
|
| | This repository hosts multiple **TorchScript-exported pretrained models** used by the **IMPACT** similarity metric for semantic medical image registration. |
| |
|
| | The IMPACT metric is described in the following preprint, currently under review: |
| |
|
| | > **IMPACT: A Generic Semantic Loss for Multimodal Medical Image Registration** |
| | > *V. Boussot, C. Hémon, J.-C. Nunes, J. Downling, S. Rouzé, C. Lafond, A. Barateau, J.-L. Dillenseger* |
| | > [arXiv:2503.24121 [cs.CV]](https://arxiv.org/abs/2503.24121) |
| |
|
| | --- |
| |
|
| | The TorchScript models provided in this repository were exported from publicly available pretrained networks. These include: |
| |
|
| | - **TotalSegmentator (TS)** — U-Net models trained for full-body anatomical segmentation |
| | - **Segment Anything 2.1 (SAM2.1)** — Foundation model for segmentation on natural images |
| | - **DINOv2** — Self-supervised vision transformer trained on diverse datasets |
| | - **Anatomix** — Transformer-based model with anatomical priors for medical images |
| |
|
| | Each model provides multiple feature extraction layers, which can be selected independently using the corresponding model l_Layers. This can be configured through the LayerMask parameter in the IMPACT configuration. |
| | |
| | In addition, the repository also includes: |
| | |
| | - **MIND** — A handcrafted Modality Independent Neighborhood Descriptor, wrapped in TorchScript |
| | |
| | --- |
| | |
| | ## 📚 Pretrained Model References |
| | |
| | | Model | Specialization | Paper / Reference | Field of View | License | |
| | |----------------|----------------------------------------|-------------------------------------------------------------|------------------------|--------------| |
| | | **MIND** | Handcrafted descriptor | [Heinrich et al., 2012](https://doi.org/10.1016/j.media.2012.05.008) | `2r + 1` | Research only | |
| | | **SAM2.1** | General segmentation (natural images) | [Ravi et al., 2023](https://arxiv.org/abs/2408.00714) | 29 | MIT | |
| | | **TS Models** | Multi-resolution CT/MRI segmentation | [Wasserthal et al., 2022](https://arxiv.org/abs/2208.05868) | `2^l + 3` | Apache 2.0 | |
| | | **Anatomix** | Anatomy-aware transformer encoder | [Dey et al., 2024](https://arxiv.org/abs/2411.02372) | Hierarchical | MIT | |
| | | **DINOv2** | Self-supervised vision transformer | [Oquab et al., 2023](https://arxiv.org/abs/2304.07193) | Global / ViT-Base | MIT | |
| | |
| | --- |
| | |
| | ### 🔍 TS Model Variants |
| | |
| | **TS Models** refer to the following TotalSegmentator-derived TorchScript models: |
| | `M258, M291, M293, M294, M295, M297, M298, M730, M731, M732, M733, M850, M851` |
| | |
| | Each model is specialized for a specific anatomical structure or resolution (e.g., 3mm / 6mm) and shares the same encoder-decoder architecture. |
| | |
| | --- |
