Convolutional Networks for Alzheimer's Classification
Read the paper here.
This work began in June 2017 as my project at the Communications Engineering Branch of the National Institutes of Health. I finished this work in November 2017, a few months after I left the NIH at the end of Summer 2017.
Using convolutional networks to classify subjects into five categories of cognitive impairment, solely from rs-fMRI data. The classes are listed below in order of increasing cognitive impairment:
- No cognitive impairment (Normal)
- Significant memory concern (SMC)
- Early Mild Cognitive Impairment (EMCI)
- Late Mild Cognitive Impairment (LMCI)
- Alzheimer's
All data used in this project is from the Alzheimer's Disease Neuroimaging Initiative. These five categories are defined more precisely in documents on the ADNI website.
This work uses a standard Inception-ResNet-v2 model for classification.
This project was fairly straightforward, but took more time than initially allocated due to the sheer size of the ADNI dataset. The three main tasks were:
- Preprocessing and sanitizing the fMRI data
- Training and testing the CNN
- Visualizing CNN results using Picasso and calculating classification metrics.
See the scripts/ folder for preprocessing and sanitization scripts, the visualize/ folder for scripts to help with Picasso visualization, and the results/ folder for a few classification metric scripts.