Non-invasive measurement of PD-L1 status and prediction of immunotherapy response using deep learning of PET/CT images
Description
1.Install python 3.5, keras 2.1.3 and Matlab 2019a (1h is needed to intall all softwares)
2.Download and extract all the folders, and the trained model (lungIO.hdf5) was in the folder named model.
3.Run GenerateNPY.m in the folder named code with Matlab to obtain all the sample .npy file from the .png file
4.Run Code- test_samplepatient.py n the folder named code with python to generate the prediction results of sample data in the folder named data.
5.Run Code- test_allpatient.py with python to generate the prediction results of all the input data in the folder named Alldata. (35 seconds are needed to run all the data)
7.Run Code- Obtainresults.m with Matlab to obtain the predicted DLS of each patient. Other test codes were also provided in this file(25 seconds are needed to obtain the final results)
Files
Alldata.zip
Files
(1.2 GB)
| Name | Size | |
|---|---|---|
|
md5:e33ac8aa344876dd7a44017295aecb6e
|
1.2 GB | Preview Download |
|
md5:28e9b52e8ada3d3879029520989a65e1
|
26.0 kB | Preview Download |
|
md5:40911e3e59ff6a9ee30396eb4dd38046
|
315.0 kB | Preview Download |
|
md5:89cb08f88e54e28ecd8267ef3c3c39fd
|
10.6 MB | Preview Download |
|
md5:25869fbd820c92fd79a96a5ebd972021
|
560.3 kB | Preview Download |