DeepStream-MRI
A complete automated analytics toolkit and workflow for AI-based brain MRI data processing, quantification, analysis and understanding.
Info/Github
- I. Automated processing streamline for mouse brain
MRI
- Structure MRI (Automated processing workflow for mouse brain
structural MRI)
- Brukertools
- AFT-Net and CU-Net (Access permissions upon request)
- Accelerated MRI and Super-resolution recontruction (Access permissions upon request)
- DL-BET: Deep Learning based Brain Extraction Tool
- Cycle Inverse Consistent TransMorph: Cycle Inverse Consistent Deformable Medical Image Registration with Transfomer (Access permissions upon request)
- CBV fMRI (Automated processing workflow for mouse brain CBV functional MRI)
- AI-CBV mapping from T2-weighted MRI
- AI-Ktrans mapping from low-dose DCE MRI
- ST-Net: Deep Learning Enables Reduced Gadolinium Dose for Contrast-Enhanced Blood-Brain Barrier Opening Quantitative
- Structure MRI (Automated processing workflow for mouse brain
structural MRI)
- II. Automated mouse brain MRS processing streamline
- III. Automated human brain MRI processing
streamline
- Structure MRI segmentation and registration
- AI-CBV mapping from T1-weighted MRI
- IV. Automated human brain MRS processing streamline
- V. Cuttlebase is a scientifictoolkit for the dwarf
cuttlefish, Sepia bandensis
- To create an MRI-based 3D brain atlas for the dwarf cuttlefish, ex vivo magnetic resonance imaging of 8 adult dwarf cuttlefish brains (4 males, 4 females) at 50 µm isotropic resolution was performed. Deep learning techniques were used to improve manually-annoted brain masks that were then used to extract the brain regions for subsequent diffeomorphic registration using the ANTs toolbox.
- VI. Statistical analyses and visualization for large neuroimaging dataset
- VII. Predicting brainAGE with MRI using deep
learning
- Regression model input: T1-weighted sMRI (to be announced)
- Regression model input: T1-weighted sMRI + AI-CBV fMRI (to be announced)
- VIII. Diagnosing Alzheimer’s disease with MRI using
deep learning
- Classification model input: T1-weighted sMRI (to be announced)
- Classification model input: T1-weighted sMRI + AI-CBV fMRI (to be announced)
- IX. Diagnosing schizophrenia with MRI using deep
learning
- Classification model input: T1-weighted sMRI (to be announced)
- Classification model input: T1-weighted sMRI + AI-CBV fMRI (to be announced)