Skip to content

eminSerin/DeepTaskGen

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

DeepTaskGen

PyTorch implementation of DeepTaskGen, a convolutional neural network designed to predict whole-brain voxel-wise individual task contrast maps from voxel-to-ROI resting-state connectomes. The architecture adapts and further extends BrainSurfCNN (Ngo et al., 2022) for volumetric representation of the brain, incorporating features such as attentional gates and Contrast-Regularized Reconstructive Loss.

Training and Evaluation

Scripts for training DeepTaskGen and validating it on three separate datasets (HCP-YA, HCP-D, and UKB) are available at https://github.com/eminSerin/deeptaskgen-paper.

Pretrained Models

Pretrained models can be found at https://github.com/eminSerin/deeptaskgen-models.

References

  1. Ngo, Gia H., et al. "Predicting individual task contrasts from resting‐state functional connectivity using a surface‐based convolutional network." NeuroImage 248 (2022): 118849.
  2. Ronneberger, Olaf, Philipp Fischer, and Thomas Brox. "U-net: Convolutional networks for biomedical image segmentation." International Conference on Medical image computing and computer-assisted intervention. Springer, Cham, 2015.
  3. Oktay, Ozan, et al. "Attention u-net: Learning where to look for the pancreas." arXiv preprint arXiv:1804.03999 (2018).

Acknowledgement

Funded by the European Union (Grant agreement No 101057429). Complementary funding was received by UK Research and Innovation (UKRI) under the UK government’s Horizon Europe funding guarantee (10131373 and 10038599) and the National Key R&D Program of Ministry of Science and Technology of China (MOST 2023YFE0199700). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union, the European Health and Digital Executive Agency (HADEA), UKRI or MOST. Neither the European Union nor HADEA nor UKRI nor MOST can be held responsible for them. 9f4d5533-ea3b-4d93-a120-9859edddb3b0_logo

About

DeepTaskGen: Volumetric-based CNN for Predicting Task Contrasts from Resting State Connectome

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages