Fundus Image Toolbox

Fundus quality prediction

A quality prediction model for fundus images (gradeable vs. ungradeable) based on an ensemble of 10 models (ResNets and EfficientNets) trained on DeepDRiD and DrimDB data. Can be just used for prediction or retrained.
Read more.

Fundus fovea and optic disc localization

A model to predict the center coordinates of the fovea and the optic disc in fundus images based on a multi-task EfficientNet trained on ADAM, REFUGE and IDRID datasets. Can be just used for prediction or retrained.
Read more.

Example predictions from the external dataset "DeepDRiD".

Fundus registration

Align a fundus photograph to another fundus photograph from the same eye using SuperRetina (Liu et al., 2022). Image registration also goes by the terms image alignment and image matching.
Read more.

Fundus vessel segmentation

Segment the blood vessels in a fundus image using an ensemble of FR-U-Nets trained on the FIVES dataset (Köhler et al., 2024).
Read more.

Fundus circle crop

Fastly crop fundus images to a circle and center it (Fu et al., 2019).
Read more.

Fundus utilities

A collection of additional utilities that can come in handy when working with fundus images.
Read more.

• ImageTorchUtils: Image manipulation based on Pytorch tensors.
• Balancing: A script to balance a torch dataset by both oversampling the minority class and undersampling the majority class from imbalanced-dataset-sampler.
• Fundus transforms: A collection of torchvision data augmentation transforms to apply to fundus images adapted from pytorch-classification.
• Get pixel mean std: A script to calculate the mean and standard deviation of the pixel values of a dataset by channel.
• Get efficientnet resnet: Getter for torchvision models with efficientnet and resnet architectures initialized with ImageNet weights.
• Lr scheduler: Get a pytorch learning rate scheduler (plus a warmup scheduler) for a given optimizer: OneCycleLR, CosineAnnealingLR, CosineAnnealingWarmRestarts.
• Multilevel 3-way split: Split a pandas dataframe into train, validation and test splits with the options to split by group (i.e. keep groups together) and stratify by label. Wrapper for multi_level_split.
• Seed everything: Set seed for reproducibility in python, numpy and torch.

Usage

The following code summarises the usage of the toolbox. See the usage_all.ipynb for a tutorial notebook and examples directory for more detailed usage examples information on the respective packages.

# Get sample images. All methods work on path(s) to image(s) or on image(s) as numpy arrays, tensors or PIL images.
fundus1, fundus2 = "path/to/fundus1.jpg", "path/to/fundus2.jpg"

import fundus_image_toolbox as fit
fundus1_cropped = fit.crop(fundus1, size=(512,512)) # > np.ndarray (512, 512, 3) uint8

import fundus_image_toolbox as fit
model, _ = fit.load_fovea_od_model(device="cuda:0")
coordinates = model.predict([fundus1, fundus2]) # > List[np.ndarray[fovea_x,fovea_y,od_x,od_y], ...]
fit.plot_coordinates([fundus1, fundus2], coordinates)

import fundus_image_toolbox as fit
ensemble = fit.load_quality_ensemble(device="cuda:0")
confs, labels = fit.ensemble_predict_quality(ensemble, [fundus1, fundus2], threshold=0.5) # > np.ndarray[conf1, conf2], np.ndarray[label1, label2]
for img, conf, label in zip([fundus1, fundus2], confs, labels):
    fit.plot_quality(img, conf, label, threshold=0.5)

import fundus_image_toolbox as fit

config = fit.get_registration_config()
# if wanted, change the config dictionary
model, matcher = fit.load_registration_model(config)

moving_image_aligned = fit.register(
    fundus1, 
    fundus2, 
    show=True, 
    show_mapping=False, 
    config=config, 
    model=model, 
    matcher=matcher
) # > np.ndarray (h_in, w_in, 3) uint8

import fundus_image_toolbox as fit
ensemble = fit.load_segmentation_ensemble(device=device)
vessel_masks = fit.ensemble_predict_segmentation(ensemble, [fundus1, fundus2], threshold=0.5, size=(512, 512)) # > np.ndarray[np.ndarray[h_in, w_in], ...] float64
fit.plot_masks([fundus1, fundus2], vessel_masks)

Installation

Install the toolbox

You can install the latest tagged version of the toolbox by running:

pip install fundus_image_toolbox

or the development version by running:

pip install git+https://github.com/berenslab/fundus_image_toolbox

Create a virtual environment

Alternatively, you can create a new conda environment and install the toolbox there:

conda create --name fundus_image_toolbox python=3.9.19 pip
conda activate fundus_image_toolbox

And then pip install fundus_image_toolbox or pip install . within the new environment.
Or create a new virtual environment including the toolbox with uv:

uv venv
source .venv/bin/activate

Contribute

You are very welcome to contribute to the toolbox. Please raise an issue or create a pull request to do so. Please feel free to contact us if you have any questions or need help via julius.gervelmeyer [at] uni-tuebingen.de.

Cite

If you use this toolbox in your research, please consider citing it:

TODO: Have a doi to cite. Currently, this work is under review.

If you use external parts of the toolbox that this toolbox provides an interface for, please consider citing the respective papers:

• Fundus registration: Liu et al., 2022
• Fundus vessel segmentation: Köhler et al., 2024
• Fundus circle crop: Fu et al., 2019

License

The toolbox is licensed under the MIT License. See the license file for more information.