| Name | medvae JSON |
| Version |
0.0.4
JSON |
| download |
| home_page | None |
| Summary | MedVAE is a family of six medical image autoencoders that can encode high-dimensional medical images into latent representations. |
| upload_time | 2025-01-31 10:08:42 |
| maintainer | None |
| docs_url | None |
| author | Stanford Machine Intelligence for Medical Imaging (MIMI) |
| requires_python | >=3.9 |
| license | MIT License Copyright (c) 2025 Stanford MIMI Lab Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. |
| keywords |
deep learning
image compression
compression
efficiency
computer aided diagnosis
medical image analysis
autoencoders
representation learning
med-vae
medvae
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# MedVAE: Efficient Automated Interpretation of Medical Images with Large-Scale Generalizable Autoencoders
[](LICENSE)
This repository contains the official PyTorch implementation for MedVAE: Efficient Automated Interpretation of Medical Images with Large-Scale Generalizable Autoencoders.
[[🤗 Hugging Face](https://huggingface.co/ashwinkumargb/MedVAE)]
<!--  -->
## 🫁 What is MedVAE?
MedVAE is a family of six large-scale, generalizable 2D and 3D variational autoencoders (VAEs) designed for medical imaging. It is trained on over one million medical images across multiple anatomical regions and modalities. MedVAE autoencoders encode medical images as downsized latent representations and decode latent representations back to high-resolution images. Across diverse tasks obtained from 20 medical image datasets, we demonstrate that utilizing MedVAE latent representations in place of high-resolution images when training downstream models can lead to efficiency benefits (up to 70x improvement in throughput) while simultaneously preserving clinically-relevant features.
## ⚡️ Installation
To install MedVAE, you can simply run:
```python
pip install medvae
```
For an editable installation, use the following commands to clone and install this repository.
```python
git clone https://github.com/StanfordMIMI/MedVAE.git
cd medvae
pip install -e .[dev]
```
## 🚀 Usage Instruction
We allow for easy-to-use CLI inference tool for compressing your high-dimensional medical images into usable latents:
```python
medvae_inference -i INPUT_FOLDER -o OUTPUT_FOLDER -model_name MED_VAE_MODEL -modality MODALITY
```
## 📎 Citation
If you find this repository useful for your work, please cite the cite the original paper:
```bibtex
@article{varma2025medvae,
title = {MedVAE: Efficient Automated Interpretation of Medical Images with Large-Scale Generalizable Autoencoders},
author = {Maya Varma, Ashwin Kumar, Rogier van der Sluijs, Sophie Ostmeier, Louis Blankemeier, Pierre Chambon, Christian Bluethgen, Jip Prince, Curtis Langlotz, Akshay Chaudhari},
year = {2025},
publisher = {Github},
journal = {Github},
howpublished = {https://github.com/StanfordMIMI/MedVAE}
}
```
This repository is powered by [Hydra](https://github.com/facebookresearch/hydra) and [HuggingFace Accelerate](https://github.com/huggingface/accelerate). Our implementation of MedVAE is inspired by prior work on diffusion models from [CompVis](https://github.com/CompVis/latent-diffusion) and [Stability AI](https://github.com/Stability-AI/stablediffusion).
Raw data
{
"_id": null,
"home_page": null,
"name": "medvae",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.9",
"maintainer_email": null,
"keywords": "deep learning, image compression, compression, efficiency, computer aided diagnosis, medical image analysis, autoencoders, representation learning, Med-VAE, medvae",
"author": "Stanford Machine Intelligence for Medical Imaging (MIMI)",
"author_email": "Maya Varma <mayavarma@cs.stanford.edu>, Ashwin Kumar <akkumar@stanford.edu>, Rogier van der Sluijs <sluijs@stanford.edu>",
"download_url": "https://files.pythonhosted.org/packages/30/14/317b33f88c6f82fc073221452708d3aaeb72c1b3893b542de4cbcd6c9b50/medvae-0.0.4.tar.gz",
"platform": null,
"description": "# MedVAE: Efficient Automated Interpretation of Medical Images with Large-Scale Generalizable Autoencoders\n[](LICENSE)\n\nThis repository contains the official PyTorch implementation for MedVAE: Efficient Automated Interpretation of Medical Images with Large-Scale Generalizable Autoencoders.\n\n[[\ud83e\udd17 Hugging Face](https://huggingface.co/ashwinkumargb/MedVAE)]\n\n<!-- ![Overview](assets/overview.png \"\") -->\n\n## \ud83e\udec1 What is MedVAE?\n\nMedVAE is a family of six large-scale, generalizable 2D and 3D variational autoencoders (VAEs) designed for medical imaging. It is trained on over one million medical images across multiple anatomical regions and modalities. MedVAE autoencoders encode medical images as downsized latent representations and decode latent representations back to high-resolution images. Across diverse tasks obtained from 20 medical image datasets, we demonstrate that utilizing MedVAE latent representations in place of high-resolution images when training downstream models can lead to efficiency benefits (up to 70x improvement in throughput) while simultaneously preserving clinically-relevant features.\n\n## \u26a1\ufe0f Installation\n\nTo install MedVAE, you can simply run:\n\n```python\npip install medvae\n```\n\nFor an editable installation, use the following commands to clone and install this repository.\n```python\ngit clone https://github.com/StanfordMIMI/MedVAE.git\ncd medvae\npip install -e .[dev]\n```\n\n## \ud83d\ude80 Usage Instruction\n\nWe allow for easy-to-use CLI inference tool for compressing your high-dimensional medical images into usable latents:\n\n```python\nmedvae_inference -i INPUT_FOLDER -o OUTPUT_FOLDER -model_name MED_VAE_MODEL -modality MODALITY\n```\n\n## \ud83d\udcce Citation\nIf you find this repository useful for your work, please cite the cite the original paper:\n\n```bibtex\n@article{varma2025medvae,\n title = {MedVAE: Efficient Automated Interpretation of Medical Images with Large-Scale Generalizable Autoencoders},\n author = {Maya Varma, Ashwin Kumar, Rogier van der Sluijs, Sophie Ostmeier, Louis Blankemeier, Pierre Chambon, Christian Bluethgen, Jip Prince, Curtis Langlotz, Akshay Chaudhari},\n year = {2025},\n publisher = {Github},\n journal = {Github},\n howpublished = {https://github.com/StanfordMIMI/MedVAE}\n}\n```\n\nThis repository is powered by [Hydra](https://github.com/facebookresearch/hydra) and [HuggingFace Accelerate](https://github.com/huggingface/accelerate). Our implementation of MedVAE is inspired by prior work on diffusion models from [CompVis](https://github.com/CompVis/latent-diffusion) and [Stability AI](https://github.com/Stability-AI/stablediffusion).\n",
"bugtrack_url": null,
"license": "MIT License Copyright (c) 2025 Stanford MIMI Lab Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the \"Software\"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED \"AS IS\", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.",
"summary": "MedVAE is a family of six medical image autoencoders that can encode high-dimensional medical images into latent representations.",
"version": "0.0.4",
"project_urls": {
"homepage": "https://github.com/StanfordMIMI/MedVAE",
"repository": "https://github.com/StanfordMIMI/MedVAE"
},
"split_keywords": [
"deep learning",
" image compression",
" compression",
" efficiency",
" computer aided diagnosis",
" medical image analysis",
" autoencoders",
" representation learning",
" med-vae",
" medvae"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "a321aa36ed0b2046f6dd1d9a1b05e62d3cef7daabf0c8076f76705ab10ea4eab",
"md5": "ca74a4f49dadbb504eb06940c3063739",
"sha256": "43ca976449dc786dc169762b7af0654ca0156ec11ebb736a19674192eed21999"
},
"downloads": -1,
"filename": "medvae-0.0.4-py3-none-any.whl",
"has_sig": false,
"md5_digest": "ca74a4f49dadbb504eb06940c3063739",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.9",
"size": 27081,
"upload_time": "2025-01-31T10:08:39",
"upload_time_iso_8601": "2025-01-31T10:08:39.354872Z",
"url": "https://files.pythonhosted.org/packages/a3/21/aa36ed0b2046f6dd1d9a1b05e62d3cef7daabf0c8076f76705ab10ea4eab/medvae-0.0.4-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "3014317b33f88c6f82fc073221452708d3aaeb72c1b3893b542de4cbcd6c9b50",
"md5": "a0525110cbfae1b7f1369d10fc734f44",
"sha256": "4cf46343cf02c689856f82394a3964aad32026e79c4d5d9ec71feb3ed301e9ff"
},
"downloads": -1,
"filename": "medvae-0.0.4.tar.gz",
"has_sig": false,
"md5_digest": "a0525110cbfae1b7f1369d10fc734f44",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.9",
"size": 23800,
"upload_time": "2025-01-31T10:08:42",
"upload_time_iso_8601": "2025-01-31T10:08:42.162606Z",
"url": "https://files.pythonhosted.org/packages/30/14/317b33f88c6f82fc073221452708d3aaeb72c1b3893b542de4cbcd6c9b50/medvae-0.0.4.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-01-31 10:08:42",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "StanfordMIMI",
"github_project": "MedVAE",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"lcname": "medvae"
}
