Superfast spherical surface processing pipeline (s3pipeline)

Python-based spherical cortical surface processing tools, including spherical mapping, resampling, interpolation, registration, parcellation, etc. It provides fast and accurate cortical surface-based data analysis using deep learning techniques.

Install

It can be installed from PyPI using:

pip install s3pipe

or using conda to directly create an environment from environment.yml

conda env create -f environment.yml
conda activate s3env

Then install pytorch from https://pytorch.org/get-started/locally/ with correct gpu/cpu and cuda choices.

Main tools

I/O vtk file. Python function for reading and writing .vtk surface file. Example code:

from s3pipe.utils.vtk import read_vtk, write_vtk

surf = read_vtk(file_name)
# Some operations to the surface 
write_vtk(surf, new_file_name)

For matlab users, please refer to this issue and this repository.

Layers provide basic spherical convolution, pooling, upsampling layers for constructing spherical convolutional neural networks.

Models provide some baseline spherical convolutional neural networks, e.g., Spherical U-Net, Spherical SegNet, Spherical VGG, Spherical ResNet, Spherical CycleGAN, etc.

Resample feature on spherical surface to standard icosahedron subdivision spheres. Example code:

from s3pipe.utils.interp_numpy import resampleSphereSurf
from s3pipe.utils.vtk import read_vtk, write_vtk
from s3pipe.utils.utils import get_sphere_template

template_163842 = get_sphere_template(163842)
data = read_vtk(file)
resampled_feat = resampleSphereSurf(data['vertices'], template_163842['vertices'], 
                                    np.concatenate((data['sulc'][:,np.newaxis], data['curv'][:,np.newaxis]), axis=1))
surf = {'vertices': template_163842['vertices'], 
         'faces': template_163842['faces'],
         'sulc': resampled_feat[:,0],
         'curv': resampled_feat[:,1]}
write_vtk(surf, file.replace('.vtk', '.resample.vtk'))

Note if you want to run it on GPU, change interp_numpy to interp_torch.

Resample label on spherical surface to standard icosahedron subdivision spheres. Example code:

from s3pipe.utils.vtk import read_vtk, write_vtk,
from s3pipe.utils.interp_numpy import resample_label
from s3pipe.utils.utils import get_sphere_template

template_163842 = get_sphere_template(163842) 
surf = read_vtk(file)
resampled_label = resample_label(surf['vertices'], template_163842['vertices'], surf['par_fs_vec'])

Cortical surface parcellation with trained models based on this package.

Cortical surface registration with trained models based on this package.

Rigid surface alignment.

Spherical surface mapping with trained models based on this package.

Surface inflation.

Surface and feature smooth.

Chcek folded triangles.

Papers

If you use this code, please cite:

Fenqiang Zhao, et.al. Spherical U-Net on Cortical Surfaces: Methods and Applications. Information Processing in Medical Imaging (IPMI), 2019.

Fenqiang Zhao, et.al. Spherical Deformable U-Net: Application to Cortical Surface Parcellation and Development Prediction. IEEE Transactions on Medical Imaging, 2021.

Fenqiang Zhao, et.al. S3Reg: Superfast Spherical Surface Registration Based on Deep Learning. IEEE Transactions on Medical Imaging, 2021.

Fenqiang Zhao, et.al. Fast spherical mapping of cortical surface meshes using deep unsupervised learning. MICCAI 2022.