An automated pipeline to detect hydrocephalus from brain MRIs
You will need Matlab R2017b or newer, Python 2.7, Tensorflow 2.1.0, and Keras 2.2.4.
After you download the zip file, unzip it, launch your Matlab, make sure you are under the root directory (i.e., you can see lib/, and the main.m file). Open the main.m, and enter the MRI to your subjects in the subjList variable at the 5th line, then run the main.m.
The program will print out the probability of hydrocephalus for each subject head MRI you entered. Also you will get the segmentation of the input head MRIs in the same folder of the MRIs.
If you use this code in your research, please cite:
Huang, Y., Moreno, R., Malani, R., Meng, A., Swinburne, N., Holodny, A., Choi, Y., Rusinek, H., Golomb, J., George, A., Parra, L., Young, R., "Deep Learning Achieves Neuroradiologist-Level Performance in Detecting Hydrocephalus Requiring Treatment", Journal of Digital Imaging (2022). https://doi.org/10.1007/s10278-022-00654-3
This work was supported in part through the NIH grants P30CA008748, R01NS095123, R01 CA247910, R01MH111896, R21NS115018, R01CA247910, R21NS115018, R01DC018589, R01MH111439, and P30AG066512-02. Support was also provided by the Memorial Sloan Kettering Cancer Center Department of Radiology.
General Public License version 3 or later. See LICENSE.md for details.
This software uses free packages from the Internet, except Matlab, which is a proprietary software by the MathWorks. You need a valid Matlab license to run this software.
The license only applies to the scripts and documentations in the root directory in this package and excludes those programs stored in the lib/ directory. The software under lib/ follow their respective licenses. This software is only intended for non-commercial use.
(c) Yu (Andy) Huang, Department of Radiology, Memorial Sloan Kettering Cancer Center
March 2022