#DONE ✓snap to grid ✓ put all results into the overleaf ✓ evaluate latest highres model ✓ train oai knees with LNCC + augmentation ✓ merge biag .vimrc ✓ train brain lncc ✓ ICON 1.0 release ✓ get OAI pull request merged ✓ merge code in "/media/data/hastings/InverseConsistency/" into master ✓ footsteps needs to log uncommitted file if that's what's being run
TODO
Cleanup todo
update OAI_analysis_2 to ICON 1.0.0
log sample registrations to tensorboard
log DICE metric to tensorboard
fix batchfunction callback- needs refactor
compress OAI models by dropping identitymaps
add identitymap stripping code to ICON
put augmentation into train.py
ICON default network
use better torch parallelism abastraction
investigate corner alignment (RIP LOL)
freeze network into asymmetricnet with fixed vectorfields
Ambitious todo
train a diffusion model
train a VQ-GAN
genius idea: registration that improves previous result: train by randomy resetting prev result
try segmenting latest kaggle challenge
try registering latest kaggle challenge
99%ICON: ICON but the worst 1% of inverse consistencies aren't penalized
mixture of gaussians ICON: interpret ICON as "registration outputs a gaussian at each pixel, ICON loss is log odds of exact inverse consistency. in that framework, rewrite to output a mixture of gaussians instead of a gaussian: perhaps can accurately model tearing?
need a synthetic dataset for tearing.
one approach to mitigate this is to create a very diverse zoo of intensity transfer functions for data augmentation, and hope that this forces the network to learn to register a larger complete set of transfer functions including MR -> CT even though that's not a strict 1-1 function Adrien's paper
Marc request todo
folds of lncc
Documentation of preprocessing- more comments
Documentation of train_batchfunction!
evaluation during training
example of how to view warped images
description of input_shape: the network will throw an error if the input images aren't this size.
download full copdgene image dataset
currently training with just LNCC for taste. Will try next with LNCC + augmentation
train lung with lin loss + augmnentation
put lin loss into losses.py- doesn't need to be embedded into GradientICON
make formal list of experiments for paper – journal article?
put OAI preprocessing into pretrained models
Create a set of images where the shapes are bright and the background dark. Create another set where it is the other way around. Train a network that gets as inputs either a random image pair from one or from the other. Will this network entirely fail if it is presented with a pair where one image is from one set and the other one from the other?
non neural finetune
high res finetune lowres model
contour based image visualization
ITK deficiencies todo
verify whether itk can be installed on an old version yet?
get itk to not segfault when asked to save a composite transform without []ifying it
get itkDisplacementFieldJacobianFilter wrapped for doubles
ICON lung todo
1 dataloader 1 augmentation 1 lambda
GradICON knee to do
Voxelmorph network with Gradient Inverse Consistency loss our lung architecture with Gradient Inverse Consistency loss
Back to Reports