FeAture Explorer (FAE), a radiomics (or medical analysis) tool that helps radiologists extract features, preprocess feature matrix, develop machine learning models (Binary Classification & Survival Analysis) with one-click, and evaluate models qualitatively and quantitatively. This project was inspired on the Radiomics, and provides a GUI with convenient process. FAE was initially developed by East China Normal University and Siemens Healthineers Ltd.

If FAE could help in your project, We appreciate that you could cite this work:

Y. Song, J. Zhang, Y. Zhang, Y. Hou, X. Yan, Y. Wang, M. Zhou, Y. Yao, G. Yang. FeAture Explorer (FAE): A tool for developing and comparing radiomics models. PLoS One. 2020. DOI: https://doi.org/10.1371/journal.pone.0237587

Welcome any issues and PR.

The Windows64 version and the Ubuntut 20.04 release could be found Google Drive or Baidu Drive. A short tutorial video with Chinese version may help.

The below modules must be installed first to make the FAE work.

- imbalanced-learn=0.6.2
- lifelines=0.27.7
- matplotlib=3.2.0
- numpy=1.21.0
- pandas=2.0.1
- pdfdocument=3.3
- pillow=7.0.0
- pycox=0.2.3
- PyQt5=5.14.1
- PyQtGraph=0.10.0
- pyradiomics=3.0
- reportlab=3.5.34
- scikit-learn=1.2.2
- scikit-image=0.18.3
- scipy=1.4.1
- seaborn=0.12.2
- statsmodels=0.11.1
- pytorch=2.0.1
- trimesh=3.9.29
- yaml=6.0

Just clone it by typing in:

git clone https://github.com/salan668/FAE.git

The .ui file has to be transferred to the .py file by pyuic manually. For example, GUI/HomePage.ui should be transferred to GUI/HomePage.py file.

• HomePage: The ui file and the Starting page for all modules.
• Feature
  • SeriesMatcher. The File matcher to help determine the image files and the ROI files for each folder.
  • GUI. The ui file and the corresponding logical files including feature extraction and feature merge.
• BC: Binary Classification Pipeline
  • DataContainer. The data structure including feature array, label, cases ID, and feature names.
  • Description. The PDF generator to describe the developed BC model.
  • FeatureAnalysis. The module of the feature pipeline, including Data Balance, Normalization, Dimension Reduction, Feature Selector, Classifier, Cross Validation and the Pipeline Structure.
  • Visualization. The common visualized plots, like ROC curve and the plot of AUC against different parameters.
  • GUI. The ui files and the corresponding logical files including pre-process, model development, and visualization.
• SA: Survival Analysis Pipeline
  • Pipeline. Similar structure to BC and different modules.
  • GUI. The ui files and the corresponding logical files including model development, and visualization.
• Plugin: Plugin manager

This project is licensed under the GPL 3.0 License

• Developer:
  • Yang Song
  • Jing Zhang
  • Guang Yang
  • Chengxiu Zhang
  • Xue-xiang Cao
• Tester:
  • Xu Yan
  • Tian-jing Zhang
  • Yi-lai Pei
  • jmtaysom
  • Zhiyong Zhang
  • Yu-dong Zhang
  • Wei Guo