Further details on the procedures including samples, image processing, neural network modeling, evaluation, and validation were published in:
Dyrba et al. (2021) Improving 3D convolutional neural network comprehensibility via interactive visualization of relevance maps: evaluation in Alzheimer’s disease. Alzheimer's research & therapy 13. DOI: 10.1186/s13195-021-00924-2.
Screenshot of the InteractiveVis app
Download this Git repository.
Make sure that all dependencies are installed in your Python environment.
Go to the parent directory DeepLearningInteractiveVis and execute:
pip install -r requirements.txt
If required, modify the header of config.py to load the correct data and CNN model.
The interactive bokeh application InteractiveVis can be run for inspecting the created relevance maps overlaid on the original input images.
Here, point the Anaconda/command line prompt to the parent directory DeepLearningInteractiveVis and then run bokeh using:
bokeh serve InteractiveVis --show
InteractiveVis UML class diagram (v4)
Select subject UML sequence diagram (v3)
Copyright (c) 2020 Martin Dyrba martin.dyrba@dzne.de, German Center for Neurodegenerative Diseases (DZNE), Rostock, Germany
This project and included source code is published under the MIT license. See LICENSE for details.