Interpretation of 3D CNNs for Brain MRI Data Classification

Introduction In this notebook we will explore how the model makes a decision on the example of simple 3D CNN classificationl model on pytorch. In the current notebook we use Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset (ADNI is a global research effort that actively supports the investigation and development of treatments that slow or stop the progression of Alzheimer's disease (AD)) , preprocessed by the Clinica software platform.

Our goal will be to study various techniques for interpreting СNN models to test the reliability of neural networks in the case of MRI brain classification.

Proceesing with this notebook you confirm, that you have granted access to ADNI 1, 2, 3, GO Study Data. And your agreement on data terms and conditions.

The data accesed from the original sourse: https://ida.loni.usc.edu/pages/access/studyData.jsp?categoryId=14&subCategoryId=30