|
|
--- |
|
|
license: cc-by-4.0 |
|
|
datasets: |
|
|
- AnonRes/OpenMind |
|
|
pipeline_tag: image-feature-extraction |
|
|
tags: |
|
|
- medical |
|
|
--- |
|
|
|
|
|
# OpenMind Benchmark 3D SSL Models |
|
|
|
|
|
> **Model from the paper**: [An OpenMind for 3D medical vision self-supervised learning](https://arxiv.org/abs/2412.17041) |
|
|
> **Pre-training codebase used to create checkpoint**: [MIC-DKFZ/nnssl](https://github.com/MIC-DKFZ/nnssl) |
|
|
> **Dataset**: [AnonRes/OpenMind](https://huggingface.co/datasets/AnonRes/OpenMind) |
|
|
> **Downstream (segmentation) fine-tuning**: [TaWald/nnUNet](https://github.com/TaWald/nnUNet) |
|
|
|
|
|
--- |
|
|
|
|
|
 |
|
|
|
|
|
## Overview |
|
|
|
|
|
This repository hosts pre-trained checkpoints from the **OpenMind** benchmark: |
|
|
📄 **An OpenMind for 3D medical vision self-supervised learning** (Wald, T., Ulrich, C., Suprijadi, J., Ziegler, S., Nohel, M., Peretzke, R., ... & Maier-Hein, K. H. (2024).) |
|
|
([arXiv:2412.17041](https://arxiv.org/abs/2412.17041)) — the first extensive benchmark study for **self-supervised learning (SSL)** on **3D medical imaging** data. |
|
|
|
|
|
Each model was pre-trained using a particular SSL method on the [OpenMind Dataset](https://huggingface.co/datasets/AnonRes/OpenMind), a large-scale, standardized collection of public brain MRI datasets. |
|
|
|
|
|
**These models are not recommended to be used as-is for feature extraction.** Instead we recommend using the downstream fine-tuning frameworks for **segmentation** and **classification** adaptation, available in the [adaptation repository](https://github.com/TaWald/nnUNet). |
|
|
*While manual download is possible, we recommend using the auto-download feature of the fine-tuning repository by providing the repository URL on Hugging Face instead of a local checkpoint path.* |
|
|
|
|
|
--- |
|
|
|
|
|
## Model Variants |
|
|
|
|
|
We release SSL checkpoints for two backbone architectures: |
|
|
|
|
|
- **ResEnc-L**: A CNN-based encoder [[a](https://arxiv.org/abs/2410.23132), [b](https://arxiv.org/abs/2404.09556)] |
|
|
- **Primus-M**: A transformer-based encoder [[Primus paper](https://arxiv.org/abs/2503.01835)] |
|
|
|
|
|
Each encoder has been pre-trained using one of the following SSL techniques: |
|
|
|
|
|
| Method | Description | |
|
|
|---------------|-------------| |
|
|
| [Volume Contrastive (VoCo)](https://arxiv.org/abs/2402.17300) | Contrastive pretraining method for 3D volumes | |
|
|
| [VolumeFusion (VF)](https://arxiv.org/abs/2306.16925) | Spatial volume fusion-based segmentation SSL method | |
|
|
| [Models Genesis (MG)](https://www.sciencedirect.com/science/article/pii/S1361841520302048) | Reconstruction and denoising based pretraining method | |
|
|
| [Masked Autoencoders (MAE)](https://openaccess.thecvf.com/content/CVPR2022/html/He_Masked_Autoencoders_Are_Scalable_Vision_Learners_CVPR_2022_paper) | Default reconstruction based pretraining method | |
|
|
| [Spark 3D (S3D)](https://arxiv.org/abs/2410.23132) | Sparse reconstruction based pretraining mehtod (CNN only) | |
|
|
| [SimMIM](https://openaccess.thecvf.com/content/CVPR2022/html/Xie_SimMIM_A_Simple_Framework_for_Masked_Image_Modeling_CVPR_2022_paper.html) | Simple masked reconstruction based pretraining method (TR only) | |
|
|
| [SwinUNETR SSL](https://arxiv.org/abs/2111.14791) | Rotation, Contrastive and Reconstruction based pre-training method. | |
|
|
| [SimCLR](https://arxiv.org/abs/2002.05709) | Transfer of 2D Contrastive learning baseline method to 3D | |
|
|
|
