This is the code for the paper: Physics-guided deep learning-based image reconstruction for Fourier-domain optical coherence tomography The code is based on Python 3.7
Load the model from the website: https://zenodo.org/records/10606447
Data preparation: The Training and validation data should have been uploaded in two folders. The data was saved in '.h5' format The code that loads the data from the .h5 file is H5Dataload.py
-
Train To train a network, run: Main_train.py Optional if a printed result is desired: nohup python train_OCT_skin_Unet_384.py >result/result.log 2>&1 & The Training and validation data should have been uploaded in two files. The model is from the paper:https://opg.optica.org/ol/fulltext.cfm?uri=ol-48-3-759&id=525607
-
Test Run: Main_test.py The test data is provided in '\Test' Change the image index in the code 'H5Dataload.py ' according to the image index for testing