Update README.md
Browse files## VascX Models: Model Ensembles for Retinal Vascular Analysis from Color Fundus Images
Model ensembles from the paper ["VascX Models: Model Ensembles for Retinal Vascular Analysis from Color Fundus Images"](https://arxiv.org/pdf/2409.16016)
### Installation
To install the entire fundus analysis pipeline including fundus preprocessing, model inference code and vascular biomarker extraction:
1. Create a conda or virtualenv virtual environment, or otherwise ensure a clean environment.
2. Install the [rtnls_inference package](https://github.com/Eyened/retinalysis-inference).
3. Download the vascx models and place them in the folder pointed by `$RTNLS_MODEL_RELEASES`.
### Usage
To speed up re-execution of vascx we recommend to run the preprocessing and segmentation steps separately:
1. Preprocessing. See [this notebook](./notebooks/0_preprocess.ipynb). This step is CPU-heavy and benefits from parallelization (see notebook).
2. Inference. See [this notebook](./notebooks/1_segment_preprocessed.ipynb). All models can be ran in a single GPU with >10GB VRAM.
