MedicalNet for classification

The MedicalNet project aggregated the dataset with diverse modalities, target organs, and pathologies to to build relatively large datasets. Based on this dataset, a series of 3D-ResNet pre-trained models and corresponding transfer-learning training code are provided.

This repository is an unofficial implementation of Tencent's Med3D model (Med3D: Transfer Learning for 3D Medical Image Analysis), originally developed for 3d segmentation tasks. It has been adapted for classification tasks using the 3D-ResNet backbone and made compatible with the Hugging Face library.

License

MedicalNet is released under the MIT License (refer to the LICENSE file for details).

Citing MedicalNet

If you use this code or pre-trained models, please cite the following:

    @article{chen2019med3d,
        title={Med3D: Transfer Learning for 3D Medical Image Analysis},
        author={Chen, Sihong and Ma, Kai and Zheng, Yefeng},
        journal={arXiv preprint arXiv:1904.00625},
        year={2019}
    }

Model Sources

Repository: https://github.com/Tencent/MedicalNet (original)
Unofficial Torch Hub Wrapper: https://github.com/Warvito/MedicalNet-models
Unofficial Huggingface Wrapper: https://github.com/JINAILAB/medicalnet3d-huggingface

How to Get Started with the Model

from transformers import AutoConfig, AutoModelForImageClassification
import torch

config = AutoConfig.from_pretrained(
    'nwirandx/medicalnet-resnet3d50-23datasets',
    trust_remote_code=True
)

# use a model from scratch
# model = AutoModelForImageClassification.from_config(
#     config,
#     trust_remote_code=True
# )

# use pretrained model
model = AutoModelForImageClassification.from_pretrained(
    'nwirandx/medicalnet-resnet3d50-23datasets',
    trust_remote_code=True
)

x = torch.randn(1, 1, 64, 64, 64)  # Example 3D volume
outputs = model(x)

MedicalNet Model Family

Original MedicalNet Series (Tencent on Hugging Face)

Unofficial Versions of the MedicalNet Classification Model Series