关于最近一些MRI-T1w结构像处理pipeline的测试

叶辰飞

3月 12, 2023 2 分钟阅读时长技巧

FastSurfer -> 全脑分割

https://deep-mi.org/research/fastsurfer/

FastSurfer is a fast and extensively validated deep-learning pipeline for the fully automated processing of structural human brain MRIs. As such, it provides FreeSurfer conform outputs, enables scalable big-data analysis and time-critical clinical applications such as structure localization during image acquisition or extraction of quantitative measures.

# 完整分割（GPU） 
docker run --gpus all -v /data/mg_data/mnt/bl_test/input:/data -v /data/mg_data/mnt/bl_test/fastsf_output:/output -v /data/freesurfer:/fs_license --rm deepmi/fastsurfer:gpu-v1.1.1 --fs_license /fs_license/freesurfer_license.txt --t1 /data/t1.nii --sid subject2 --sd /output --parallel 

# DKT Lookup table 
31和63作为脉络丛会被分割出来，应该当做LV，对应4和43 
77作为WMH，没有区分左右，不便合并

SynthSeg -> 全脑分割，计算ICV

https://github.com/BBillot/SynthSeg https://surfer.nmr.mgh.harvard.edu/fswiki/SynthSeg

SynthSeg 可不依赖GPU，实测CPU多线程计算约2分钟内完成

# 分割 （完整） 
mri_synthseg --i /dataio/t1.nii --parc --robust --vol /dataio/vol.csv --qc /dataio/qc.csv --threads 172 --o /dataio/t1_seg.nii.gz 
# 注意 --resample如果输入本身是1mm iso，则不会生成resample图像

SynthStrip -> 剥头皮

https://surfer.nmr.mgh.harvard.edu/docs/synthstrip/

SynthStrip is a skull-stripping tool that extracts brain signal from a landscape of image types, ranging across imaging modality, contrast, resolution, and subject population. It leverages a deep learning strategy that synthesizes arbitrary training images from segmentation maps to optimize a robust model agnostic to acquisition specifics. https://doi.org/10.1016/j.neuroimage.2022.119474

# 自动去头皮（多模态泛化性优秀），1分钟以内 
mri_synthstrip -i /dataio/t1.nii -o /dataio/t1_bet.nii -m /dataio/t1_mask.nii

sbTIV-> 计算ICV

https://freesurfer.net/fswiki/Samseg https://freesurfer.net/fswiki/sbTIV Total intracranial volume (TIV/ICV) is an important covariate for volumetric analyses of the brain and brain regions. It is commonly used to correct for head size variation (i.e., ’normalize’ hippocampal volume size). The gold-standard method is manual delineation of T2 scans. Freesurfer currently provides the eTIV measure, described here. It has been shown to be a robust covariate.

## 计算sbTIV (segmentation-based TIV)，大约5-10分钟
run_samseg --input /home/username/data/t1.nii --output /home/username/data/samsegOutput/ --threads 8

关于FreeSurfer图谱

DK图谱lookuptable https://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/AnatomicalROI/FreeSurferColorLUT https://www.cis.jhu.edu/~parky/MRN/Desikan%20Region%20Labels%20and%20Descriptions.pdf
DKT 图谱： https://mindboggle.readthedocs.io/en/latest/labels.html
HCPMMP图谱： https://neuroimaging-core-docs.readthedocs.io/en/latest/pages/atlases.html

pipeline docker