This repository contains the official PyTorch implementation of the following paper:
Vessel-Promoted OCT to OCTA Image Translation by Heuristic Contextual Constraints
Shuhan Li, Dong Zhang, Xiaomeng Li, Chubin Ou, Lin An, Yanwu Xu, Kwang-Ting Cheng
https://arxiv.org/abs/2303.06807
Optical Coherence Tomography Angiography (OCTA) has become increasingly vital in the clinical screening of fundus diseases due to its ability to capture accurate 3D imaging of blood vessels in a non-contact scanning manner. However, the acquisition of OCTA images remains challenging due to the requirement of exclusive sensors and expensive devices. In this paper, we propose a novel framework, TransPro, that translates 3D Optical Coherence Tomography (OCT) images into exclusive 3D OCTA images using an image translation pattern. Our main objective is to address two issues in existing image translation baselines, namely, the aimlessness in the translation process and incompleteness of the translated object. The former refers to the overall quality of the translated OCTA images being satisfactory, but the retinal vascular quality being low. The latter refers to incomplete objects in translated OCTA images due to the lack of global contexts. TransPro merges a 2D retinal vascular segmentation model and a 2D OCTA image translation model into a 3D image translation baseline for the 2D projection map projected by the translated OCTA images. The 2D retinal vascular segmentation model can improve attention to the retinal vascular, while the 2D OCTA image translation model introduces beneficial heuristic contextual information. Extensive experimental results on two challenging datasets demonstrate that TransPro can consistently outperform existing approaches with minimal computational overhead during training and none during testing.
- Create conda environment and activate it:
conda create -n octa python=3.6
conda activate octa
- Clone this repo:
git clone https://github.com/ustlsh/TransPro
cd TransPro
- Install requirements:
pip install -r requirements.txt
You can download OCTA-3M and OCTA-6M from OCTA-500
You can download weights from: VPG and HCG
Move them into the "pretrain-weights" folder.
- To view training results and loss plots, please run:
python -m visdom.server -p 6031
and click the URL http://localhost:6031
- To train the model, please run:
python train3d.py --dataroot ./octa-500/OCT2OCTA3M_3D --name transpro_3M --model TransPro --netG unet_256 --direction AtoB --lambda_A 10 --lambda_C 5 --dataset_mode alignedoct2octa3d --norm batch --pool_size 0 --load_size 304 --input_nc 1 --output_nc 1 --display_port 6031 --gpu_ids 0 --no_flip
- To test the model, please run:
python test3d.py --dataroot ./octa-500/OCT2OCTA3M_3D --name transpro_3M --test_name transpro_3M --model TransPro --netG unet_256 --direction AtoB --lambda_A 10 --lambda_C 5 --dataset_mode alignedoct2octa3d --norm batch --input_nc 1 --output_nc 1 --gpu_ids 0 --num_test 15200 --which_epoch 164 --load_iter 164
If our paper is useful for your research, please cite:
@article{li2023vesselpromoted,
title={Vessel-Promoted OCT to OCTA Image Translation by Heuristic Contextual Constraints},
author={Shuhan Li and Dong Zhang and Xiaomeng Li and Chubin Ou and Lin An and Yanwu Xu and Kwang-Ting Cheng},
journal={arXiv},
year={2023}
}