Dataloader that serves MRI images from a mogodb.
The main idea is to keep MRI images and corresponding training labels
for segmentation tasks in a mongo database. However, each 3D MRI
tensor even in 8 bit precision is 16Mb.
mongo's records cannot be
larger than this limit and we need to also store the labels of the
same dimensions. mindfultensors fetches and aggregates each
tensor stored across multiple records, together with corresponding labels
either for gray and white matter, 104 regions atlas, volume of each of 104 ROIs, or a 50 region
atlas.
The package is on pypy and the simplest way to install it is
pip install mindfultensors
However, to tinker with it you can also clone the repo:
git clone git@github.com:neuroneural/mindfultensors.git
Then change directory to the newly cloned repository:
cd mindfultensors
And install locally by
pip intall -e .
A detailed example of how to create a dataloader using provided
dataset class and the corresponding tools is in
scripts/usage_example.py
Do not forget to move the batches to the GPU once obtained.